Sei sulla pagina 1di 321

A 1d8 8ud

K C8C8tC|
uu8 1u
A uC1CD1
N C8t
L 881CtD
b1ud1C8
5
JOHN L. HA YES
A MANUAL OF SUMERIAN GRAMMAR
AND TEXTS
@LDD
tLLLDb
M81bu
1
AIS AND RESEARCH TOOLS IN ANCIENT NEAR EASTERN STUDIES
Editor: Giorgio Buccelati
This is an introductory pedagogical grammar, designed for readers with
no previous knowledge of Sumerian or its writing system, to be used either with
or without a teacher. It includes a general description of the language and its
writing system, and a series of 77 lessons. Each lesson includes: sign-list and
vocabulary; cuneiform text(s); transliteration, transcription, and translation;
linguistic commentary. The texts used are royal inscriptions of the Ur III period,
presented in photograph or autograph. A certain amount of historical, archaeological,
and cultural background is also included. While primarily meant for students
of Mesopotamia who already are familiar with Akkadian, it is also designed for
students of West-Semitic, who may know no Akkadian. For this latter audience,
emphasis has been placed on transliteration and transcription, to enable the
Manual to be used without learning the cuneiform signs.
Copyright (c) 1990 by Undena Publications
All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any
means, electronic or mechanical, i ncluding photocopy, recording, or any information storage and
retrieval system, without permission in writing from the publisher.
Library of Congress Card Catalog Number: 89-51971
ISBN 0-89003-198-3 (cloth) /0-89003-197-5 (paper)
Issued under the auspices of I1MAS-The International Institute for Mesopotamian Area Studies
LDDR tLLLR1Db t.. 7 79, MRLL LRLtDU 90265
PREFACE
Anyone who has ever tied to lea or to teach Sumerian faces a diffcult task. First of
all, knowledge of Sumerian is still at an imperfect stage, with fundamental questions yet to
be resolved. Second, there is a lack of both scholarly and pedagogical tools. Although a
recent descriptive grammar exists, there is no up-to-date sign list or dictionary, and there is
no text-book of any kind. This situation makes it diffcult for both student and teacher, and
makes it virtually impossible for someone to learn Sumerian without a teacher.
The aim of this book is to help alleviate this situation. It is a textbook of the Sumerian
language, based on the royal inscriptions of the Ur III period. It is self-contained, so that it
will be of use to students with or without a teacher. It includes a general description of the
Sumerian language and its writing system, and then a series of graduated lessons. Each
lesson contains: sign-list and vocabulary; notes on selected vocabulary; text(s) in cunei
form, either photograph or autograph; transliteration, transcription, and translation; line-by
line commentary on the text. Each lesson concludes with discussions, arranged the
matically, of grammatical issues raised by the text, and of the meaning, function, and
historical context of the text. Later lessons also include supplementary texts for review and
practice, with no new vocabulary or grammar. In each lesson the grammar has generally
been presented inductively from the texts. Finally, there are several appendices, some
teating more general topics, and some serving as reference; the last of these is an index to
grammatical (and other) points.
This book has been designed for a one-semester, three-hour per week class. It can
serve as an introduction to the language for students who will not pursue their study of
Sumerian any further, but it will also prepare students for more advanced work.
Two possible audiences are envisaged. The frst is composed of those students who
are comfortable in Akkadian, and who wish to learn Sumerian principally because of their
interest in Mesopotamia. The second is composed of those students who are more
comfortable in West-Semitic, and who wish to learn Sumerian principally because of their
interest in Ebla. The latter audience will either not have studied Akkadian at all, or will
have studied it at some time in the distant past, and may have forgotten much. A certain
amount of material for this latter audience is included which will already be known to those
who are familiar with Akkadian. Throughout, a knowledge of basic linguistic terms and
concepts has been assumed. Since the learning of cuneiform signs often seems like an
onerous chore for those students primarily interested in West-Semitic, the book has been
designed with suffcient emphasis on transliteration and transcription to allow it to be used
without leaing the signs.
This book is based on the language of the royal inscriptions of the Ur III period. It is
thus a grammar solely of the written form of the language. It attempts to be purely
synchronic, avoiding a mixture of synchronic and diachronic levels. At the same time,
areas of disagreement about the language are pointed out. Some stress has been placed on
the methodological principles involved in studying a language like Sumerian. Since many
of the problems in understanding Sumerian phonology, morphology, and even syntax are
l
iv Preface
rooted in difculties with the script, a certain emphasis has been placed on the nature of the
Sumeriim writing system.
In order to give an idea of the context in which the texts are roted, some
archaeological, historical, and cultural informaton is included. Similarly, typological
observations about the Sumerian language have been pointed out, to show that there are
other languages which work in ways similar to Sumerian.
Because of the limited subject-matter of the texts which are used here, not all features
of the language are encountered. Some of these features are touched upon in Lesson 23,
where some alternative views of Sumerian grammar are sketched. Appendix 5 discusses
the ways by which students, including those working alone, can deepen their understanding
of Sumerian. This book will be followed by a second volume, consisting of heavily
annotated extacts from Inanna
'
s Descent. The reading of a major literary text will into
duce students to a number of problems not encountered in reading the rather stereotyped
texts used in this book.
Appendix 4 is a basic bibliography of the most important and interesting books and
aticles on Sumerian. In order for students to become become acquainted with the names of
some of the scholars in the feld, a number of moder-day Assyriologists and Sumerolo
gists are quoted throughout the book; all works so quoted are listed in Appendix 4
The genesis of this book goes back to my teaching of Sumerian at the University of
California at Los Angeles. It is a pleasure to thank those who have helped out along the
way. Thorkild Jacobsen was my frst teacher of Sumerian; his infuence can easily be seen
throughout the book. Saa Denning-Bolle graciously drew the cuneiform signs used in the
sign-lists and those scattered throughout the book; I am especially grateful to her. Babara
De Marco made a number of useful stylistic observations, and helped in the overall
stucture. Several individuals read earlier gestations; I would especially like to thank Daniel
Foxvog, Samuel Greengus, and Stephen Lieberman. Other individuals read certain sec
tions; I thank Denise Schmandt-Besserat and Russell Schuh. James Platt, who studied
from this bok, made a number of suggestions. Christopher Walker helped me attain
access to a number of photos fom the British Museum. Giorgio Buccellati helped in many
ways, fom the initial conception to the fnal product. And, I would like to thank the staff at
Undena Publications, especially Frank Comparato and Paticia Oliansky. Faults remaining
are my own; I would be very grateful to hea frm readers with suggestions for revisions.
I would like to dedicate this book to my mother, for her support and encouragement
over all the years.
CONTENTS
PREFACE . . . . . . . . .
INTRODUCTION
Importance of S umerian . . . . .
Diffculties in studying Sumerian . .
Historical background and texts used
PART ONE: T SUMERIAN LANGUAGE
CLASSIICATION . .
Linguistic affliation
Dialects . . . . .
Typological characteristics
Ergativity . .
Agglutination .
WRITING SYSTEM.
Exteral characteristics
Original nature . .
Interal principles
Transliteration
Transcription
PHONOLOGY
Problems .
Vowels . .
Consonants
Other features
PART TWO: LESSONS IN SUMRIAN GRAMMAR
Lesson 1
Lesson 2
Lesson 3
Lesson 4
Lesson 5
N
i
1
1
2
5
5
5
7
7
1 0
1 1
1 1
1 2
1 3
14
16
1 8
1 8
19
20
23
25
47
57
65
73
vi Contents
Lesson 6 79
Lesson 7 85
Lesson 8 95
Lesson 9 1 01
Lesson 1 0 . 1 09
Lesson 1 1 . 1 17
Lesson 1 2 . 1 29
Lesson 1 3 . 147
Lesson 1 4 . 157
Lesson 15 . 1 65
Lesson 1 6 . 1 81
Lesson 17 . 1 91
Lesson 1 8 . 205
Lesson 1 9 . 21 1
Lesson 20 . 223
Lesson 21 . 23 1
Lesson 22 . 245
Lesson 23 . 257
PART THREE: APPENDICES
Appendix 1 : History . 265
Appendix 2: Mesopotamian sources 273
Appendix 3: Glossary 283
Appendix 4: Bibliography 291
Abbreviations 291
Works cited 291
Concordance of texts 299
Appendix 5: Further work 305
Appendix 6: Topical index 309
Contents vii
TEXTS and ILLUSTRATIONS
Text 1 : brick ofUr-Nammu -autograph 29
Text 2: brick of Ur-Nammu -photograph 47
Text 2: brick ofUr-Nammu -autograph . 49
Text 3a: brick ofUr-Nammu -photograph 57
Text 3a: brick ofUr-Nammu-autograph 59
Text 3b: brick ofUr-Nammu -autograph 60
Text 3c: brick of Ur-N ammu -autograph . 63
Text 4: votive bowl ofUr-Nammu -autograph 66
Text 4a: brick ofUr-Nammu -autograph 72
Text 5: cone ofUr-Nammu -autograph 75
Text 6: brick ofUr-Nammu -autograph 81
Text 7a: foundation tablet of Ur-N ammu -photograph 86
Text 7b: brick of Ur-N ammu -autograph . 87
Text 7c: brick ofUr-Nammu -autograph 93
Text 8: door socket ofUr-Nammu -autograph 96
Text 8a: brick of Ur-N ammu -autograph 99
Text 9: brick ofUr-Nammu -autograph 1 02
Text 1 0: weight of Shulgi -autograph 1 10
Text lOa: brick ofUr-Nammu -autograph 1 1 6
Text 1 1 : wig of Shulgi -autograph 1 20
Text l 1 a: brick of Ur-Nammu -photograph 1 28
Text 1 2: vase of Ur-Nammu -autograph 1 31
Text 1 2a: door socket of Shulgi -autograph 144
Text 1 3a: brick of Amar-Sin -autograph . 1 49
Text 1 3b: brck of Amar-Sin -autograph . 1 50
Text 1 3c: brick of Amar-Sin -autograph . 1 55
Text 14: brick of Amar-Sin -autograph . 1 59
Text 14a: cone ofUr-Nammu -autograph 1 64
Text 1 5: pedestal of Amar-Sin -autograph 1 67- 1 68
Text 1 6: door socket of Amar-Sin -autograph 1 83
Text 1 6a: bead of Shulgi -autograph 1 89
Text 17: door socket of Amar-Sin -photograph 1 91
Text 1 7: door socket of Amar-Sin -autograph 1 94
Text 1 8: door socket of Shu-Sin -autograph 206
Text 1 8a: bead of Shulgi -autograph 210
Text 19: door socket of Shu-Sin -autograph 214
Text 1 9a: amulet of Amar-Sin -autograph 221
Text 20: brick of Shu-Sin -autograph 226
Text 21 a: weight of Shu-Sin -photograph 231
Text 21b: seal of Ibbi -Sin -autograph . 234
viii Contents
Text 21 c:
Text 21 d:
Text 22:
seal of lbbi-Sin -autograph . . . . . . .
weight of Shulgi -photograph and autograph
seal of Shulgi -photograph . . . . . . .
Text 22:
Text 22a:
seal of Shulgi -autograph . . . . . . .
seal ofUr-Nammu -photograph and autograph
reconstuction of ziggurat ofUr-Nammu
remains of ziggurat ofUr-Nammu
Stela ofUr-Nammu
Old Akkadian brick-stamps
fgurine of Ur-Nammu . .
reconstruction of ziggurat of Nabonidus
duck weight . . . . . . .
door socket of Inanna Temple
Neo-Baby Ionian pedestal
Old Akkadian seal . . . . .
241
243
245
249
255
42
42
43
53
92
107
1 1 2
145
179
236
IODUCION
fponnce of Sweran
For students of Mesopotamia, the need to study Sumerian is obvious. Alongside
Akkadian, Sumerian is of prime importance for reconstucting many aspects of
Mesopotamian history and culture. However, a knowledge of Sumerian is also useful for
those students primarily interested in Semitic linguistics, and for those interested in biblical
studies.
For Semitsts, Sumerian is of importance because of its perasive infuence upon
Akkadian -infuence upon the phonology, morphology, syntax, and lexicon. Only through
a knowledge of Sumerian can one differntiate between features of Akkadian which are a
product of its Semitic ancesty, and those which have arisen secondarily under the infuence
of Sumerian.
Even though Eblaite has only been known for a sport while, it i s clear that its study
will have a profound effect on Semitic linguistics. However, the majority of the texts found
at Ebla are written in pure Sumerian, not in Eblaite. The remaining texts, although written
in the Eblaite language, are couched in a Sumerian writing system which obscures many of
the actual Eblaite fors. This means that a knowledge of Sumerian, especially a thorough
understanding of the principles underlying the Sumerian writing system, is of importance
for research in Eblaite.
Diffculties in studying Sweran
Sumerian is not as well understood as is Akkadian; a number of features in the
morphology and in the syntax are not clear. Although there has been considerable
linguistic progress in the last two decades, enough still remains unsure so that scholars
often have widely divergent views about Sumerian. Some of the reasons for these
difculties are summarized here; they will be discussed in more detail in the course of this
book.
(1) Sumerian is not genetically related to any other known language, living or dead.
By contast, it was discoverd early-on that Akkadian was a Semitic language. This genetic
relationship aided early scholars in their reconstuction of Akkadian grammar and
vocabulary. But in the case of Sumerian, there is no such help available.
(2) The writing system of Sumerian only imperfectly mirrors the spoken language; it
does not indicate all the grammatical features which are known to have existed in the
spoken language. This schematic nature of the script makes it very difcult to reconstuct
the morphology.
(3) There are many instances of sentences which seem to differ only slightly in their
morphology or syntax. But with no comparative evidence, and with no native speakers to
turn to, it is diffcult to determine what these differences in morhology and syntax may
mean. There are undoubtedly many nuances of meaning which cannot be determined at all.
It has been remarked by Igor Diakonoff, "It is a joke well known among
2 Mau of Sweria
Assyriologists that there are as many Sumerian languages as there are Sumerologists"
(1976:99). Similarly, ThorkildJacobsen has recently said:
Knowledge of Sumerian is still in a rudimentary, experimental stage where
scholars differ on essential points, so that tanslations, even by highly
competent scholars, may diverge so much that one would never guess that
they rendered the same text. ... Scholars have not yet been able to agree on
basic grdar and its restaints (1987:xv).
I certain ways, however, it is actually easier to study Sumerian than it is to study, for
example, Akkadian. This is because Sumerian does not have (at least, there is not visible) a
great deal of "morphology"; there are not a large number of grmmatcal forms to lea.
There is nothing like the weak-verb systems of Akkadian and Hebrew, which require a
great deal of sheer memorization. Rather, many students fnd the diffcultes to be more
conceptual in nature: the language works in ways different than English, or other
languages which students are likely to have been exposed to. It is sometimes diffcult to
understand some of these principles, and even more difcult to obsere these principles in
action.
Hstorcal backgound and texts used
The texts utilized here are all royal inscriptions of the Ur II Dynasty (approximately
2112-2004 BC), sometimes referred to as the Neo-Sumerian Dynasty. It grew out of the
vacuum left by the collapse of the Dynasty of Akkad, which had been ruled by Akkadian
speaking kings of Semitic stock (approximately 23342193 BC).
The Ur II Dynasty was founded by Ur-Nd u, who ruled in the city of Ur from
about 2112 to 2095. He had previously been goveror of Ur under the suzerainty of the
king of Uruk, Utu-Hengal; he may have been a relative of the latter. At some point he
declared himself independent. During his rule, and especially during the rule of his son
Shulgi, the territory contolled by Ur expanded, until it reached most of the area previously
contolled by the rulers of Akkad, that is, most of cental and souther Mesopotamia. After
three more descendants of Ur-Nammu, the dynasty collapsed in 2004, partially due to
pressures fom the intusion of nomadic, Semitic-speaking tibes. Thus, the Ur II period
lasted a little more than a century; with the fall of Ur, Sumerian civilization, for all intents
and purposes, also fell.
Ur II was a period of relative calm and stability in much of Mesopotamia. Because of
the blooming of Sumerian art and literture, which had been somewhat submerged under
the Semitic dynasty of Akkad, this period is often called the "Sumerian Renaissance".
Towns were fortifed, many temples were rebuilt, and canals were dredged; tade with
various foreign counties fourished.
The city of Ur itself, the capital of the Ur II Dynasty, was primaily excavated by Sir
Leonard Woolley, perhaps the most famous of all Near Easter archaeologists. The
principal results were published by him and others in a series entitled Ur Excavations. Ten
volumes have appeared: Volume I in 1929, and Volume VI in 1976 (Volume X appeared
in 1951). Woolley popularized his results in a one-volume work entitled Ur of the
Intoduction 3
Chaldees (1929). After Wolley
'
s death, P.R.S. Morey revised and updated the work; it
appeared as Ur 'of the Chaldees
'
(1982). This is a readable and interesting description of
the city at different historical periods.
Many Ur III texts have been presered. The vast majority ar economic and
admnistrative; these number in the tens of thousands. Unfortunately, there are very few
texts of what might be called a "historical" nature. There is much that is not known about
such matters as Ur-Nammu
'
s rise to power, the interal politics of the Ur II Dynasty, or
even the physical extent of the Ur III "Empire"; C. J. Gadd refers to the "tantalizing want
of information due to the singular unwillingness of the age to record even the tiumphs,
much less the failures, of its kings" (1971 :617).
Some original literary texts are also preserved from this period, as well as older works
now committed to writing. Jacobsen says that the kings ofUr Il, especially Shulgi,
were much concered to preserve extant older literary works and to
encourage the creation of new ones. The court background of these works is
unmistakable . . . . A major portion of Sumerian Literatur as we have it taces
back to the court of the kings of the Third Dynasty of Ur, where it was
composed and performed by the royal bads (1987 :xii, 277).
The royal inscriptions of the Ur III kings have been the object of study by W. W.
Hallo. According to Hallo
'
s definition, royal inscriptions are texts which "were dedicated
either by, or to, or on behalf of the king" (1962:1). Hallo catalogued these texts, providing
a standard system of reference. He also studied the different sub-types of ryal
inscriptions, categorizing them according to their function and according to their form.
These texts range in difculty, from quite simple to very complex. They also contain a
high degree of formulaity; many of the epithets of the king, for example, occur in a large
number of the inscriptions. Even the phrasing of the verbal expressions is rather fxed.
Since the genre of royal inscriptions existed both before and after the time of Ur III (in
Sumerian and in Akkadian), a knowledge of the Ur III texts gives immediate access to
other similar texts.
There has been much recent discussion about when Sumerian ceased to be a spoken
language. This is not an easy question to answer; there are both historical issues and issues
of general linguistics to resolve. (The subject is further discussed in Appendix 1.) Most
Sumerologists would say that Sumerian was a living spoken language in Sumer during the
Ur III period, although some would say that it was already starting to die out during the
latter part of this period. A minority would say that spoken Sumerian was either pretty far
on its road to extinction, or might even have ceased to be a spoken language by the end of
the Ur III period. Even the proponents of this view, however, would admit that the
language of the Ur III royal inscriptions is "good" Sumerian, unlike some Sumerian of later
periods.
PAT ONE : TSUME IAN LANGUAGE
CLAS S ICATON
Lingistic afliation
Sumerian appears to be what is called a language-isolate, that is, it has no genetic
connecton with any known language, living or dead. Attempts have been made to link
Sumerian with many different languages -the most popular have been Hungarian, Turkish,
Caucasoid, Dravidian, and the Indus Valley language(s) - but none of these has found
general acceptance. Such attempts have usually been based on surface-level resemblances
with languages which are typologically similar.
A. Leo Oppenheim has pointed out:
The fact that Sumerian is a complicated though very well understood
language which cannot be linked to any other known language has created
during the past hundred years a large literature attempting to relate Sumerian
to practically all languages between Polynesia and Africa. The authors of
such studies unfailingly "prove" that either their own language or a language
in which they happen to be interested is related to ancient Sumerian
( 1 971 : 219).
Sir Gerard Clauson has summed this up: "Sumerian . . . has every appearance of being
a 'loner', in spite of numerous attempts to foist relatives upon it, some grotesquely
improbable"( 1 973:38).
The possibility that a connection might be found with some other language is slim.
Any related languages have probably died off without leaving any written records. The
original homeland of the Sumerians is unknown, so it is not even clear where its possible
linguistic relatives might be located. Wherever such a homeland might be, it was probably
not in an area where writing developed very early.
Dialects
The Sumerians referred to their own language by a term often tansliterated as: eme
gi 15. The value of the second sign is not sure, and so the term is variously tansliterated as
eme-gi
7
eme-l etc. , especially in older secondary literature. ere means "tongue" in
Sumerian. The meaning of gi
l
S
is unsure. Older scholars thought that it meant "Sumer";
in that case, the term would mean "language of Sumer". More recently it has been argued
that the term means something like "noble, prince"; ere-gi
l
S
would then mean "the noble
language". Because of the uncertainties in reading this word, the ter "Main Dialect" is
often used instead.
There is also a "dialect" called ere-sal. The meaning of the second element of the
name is uncertain; it may mean "fne, thin". The "status" of this dialect is also uncertain. It
has traditionally been called a "women's language", because it appears in literary texts of the
Old Babylonian period, used by women when speaking to other women. For example, in
the myth "Inanna's Descent to the Netherworld", when Inanna speaks to her aide Nin-
b
6
Mau of Suera
Shubur, she does so in Emesal. There is no consistency in this usage; in other texts Inanna
may speak in Main Dialect. Moreover, in texts of the later Old Babylonian period Emesal
is also used for specifc genres of text. Certain kinds of lamentations are always written in
Emesal, even though recited by male priests. (Texts in some of these genres were
preserved and even composed in schools for a thousand years after Sumerian had ceased to
be a spoken language.) This use by men makes it difficult to determine exactly what
Emesal is, and whether or not it should be classifed as a "dialect".
Emesal is well-attested fom the beginning of the Old Babylonian period on.
However, there appear to be at least one or two Emesal forms in the Gudea texts, and there
has been a recent attempt to see Emesal forms in a group of texts written in an unusual
orthography fom Tell Abu Salabikh (approximately 2600 BC).
Emesal differs fom Main Dialect in phonology and in the lexicon, but not apparently
in morphology. In phonology, the Emesal fors often appear to be older. For example,
the word for "lord" in Main Dialect is len/, in Emesal lumun/. It is difficult to say exactly
what the more original form was; it may have been something like */ewenl or */uwun/.
In any case, the Emesal form appears more conservative than the Main Dialect form.
According to other scholars, however, Emesal forms are linguistically the more innovatve;
Emesal forms result from consonants being shifted to a more fonted or to a higher place of
articulation. For example, Main Dialect Igl Emesal Ib/; Main Dialect Idl Emesal Iz/,
etc. But there are several exceptions to these general principles, and there are a number of
details of Emesal phonology which are not clear. As an example from the lexicon, the
Main Dialect word for the interogative "what?" is lana/; the Emesal form is /tal. These
are apparently two etymologically distinct words.
It has been claimed that Emesal shares certain characteristics of "women's languages"
which occur elsewhere in the world. In particular, women's languages are said to differ
from "standard" dialects in phonology - the women's dialect being more conservative than
the standard dialect - and in the lexicon. More work needs to be done in defning the
characteristcs of Emesal, and in comparing Emesal with other women's languages.
Not much is known about geographical variation within Sumerian. The extent of the
Sumerian-speaking area i

unsur; Sumerian texts are presered from only a rather limited
area. Moreover, the nature of the Sumerian writing system makes it diffcult to see such
vaatlOn. Only taces can be found, particularly in the later periods. There was
undoubtedly more dialectal variation present than the writing system allows us to see.
Similarly, although Sumerian was spoken over a long period of time, there does not
appear to be much variation before the Old Babylonian period. More differentiaton is
noticeable in post-Old Babylonian periods, when Sumerian was no longer a spoken
language. But here the differences may refect the practices of different scribal schools and
scribal centers, and not differences which were originally in spoken Sumerian.
There are occasional references in late Sumerian texts to what are apparently
specialized languages, or jargons of particular occupations. For example, there are passing
references to eme-utula, "the language of shepherds", and to eme-ma-Iah
4
-' "the language
of sailors". It is hard to say what these dialects or jargons were like. Similarly, there are
only passing references to what may be some kind of "literary dialects": ere-gal, "great
Classifcation 7
language", eme-sukud, "high language", etc. It is not known what these designations
mean.
Tyological chaacteristcs
There are several ways in which Sumerian works differently than the Semitic or Indo
European languages. Consider the Akkadian sentence, "The king went":
(1) arum
king-NOM
ili
VERB
Now, consider the Akkadian sentence, "The king built the house":
(2) Sarum
king-NOM
bItam Ipu
house-A CC VERB
I Akkadian, "king" is the subject in both sentences: It is the subject of an intansitive
verb in sentence (1), and the subject of a tansitive verb in sentence (2). Therefore, in both
sentences it is put into the nominatve case, Sarrum. In sentence (2), "house" is the direct
object of a tansitive verb, and so it is put into the accusative case, bItam.
Languages in which the subject of a tansitive verb and the subject bf an intansitive
verb are marked one way (called the "nominative" case), and the direct object is marked a
different way (called the "accusative" case), are often called "accusative" languages (or
"nominative-accusative"languages) .
Sumerian, on the other hand, is what is called an "ergative" language. In an ergative
language, what we consider to be the subject of a tansitive verb is marked by the "ergative"
case. But, what we consider to be the subject of an intansitive verb, and what we
consider to be the direct object of a tansitive verb, are both marked by the "absolute" case.
I some ergative languages the ending for the ergative case, and the ending for the
absolute case, may look completely different. I other ergative languages, the ergatve case
will have one marking, but the absolute case will be unmarked. ("Unmarked" can also be
understood as "marked by zero". This can be symbolized by "zero": 0.) In other
languages, there is no case-marking on any of the nouns; rather, ergativity is refected in the
way that certain elements within the verb cross-reference the case relationships.
I Sumerian, sentences (1) and (2) would be expressed as follows (Here and
elsewhere, a period is used to separate morhemes; the verb forms have been slightly
simplifed) :
(3) luga1.0 i.gin
king-ABS VERB
(4) luga1.e e.0 mu.n.du
king-ERG house-ABS VERB
8 Mau of Sueria
I (3), the subject of the intansitive verb is marked by .0, the absolute case-marker.
In (4), the subject of the tansitive verb is marked by .e, the ergatve case-marker, while the
direct object is marked by .0, the absolute case-marker. This fts the defnition of an
ergative language: The subject of a tansitive verb is marked one way (in Sumerian, by .e),
while the subject of an intransitive verb, and the direct object of a transitive verb, are
marked a different way (in Sumerian, by .0).
Ergativity is a different way of marking the primary participants in a sentence. In an
accusative language, the subject of a tansitive verb and the subject of an intansitive verb
fall into one grammatcal category; in an ergative language, the subject of an intansitive
verb and the object of a tansitive verb fall into one grammatical category. Consider the two
English sentences, "The ball rolled down the hill", and "The boy rolled the ball down the
hill". I English, "ball" in the frst sentence is the subject, but in the second sentence it's the
direct object. Yet in each case, it's the ball that is rolling down the hill. In an ergative
language, "ball" would be in the absolute case in both the frst and second sentences, and
"boy" would be in the ergative case in the second sentence. In this example, an ergative
language seems to capture our intuitions about the role of the ball in these two sentences
better than does our accusative language.
I the above discussion, the terms "subject" and "object" were used. However, it is
imprecise (and unjustifed on theoretcal grounds) to use these two terms when talking
about an ergative language. Most linguists prefer to use the term "agent" to refer to the
subject of the tansitive verb (marked by the ergative case), and the term "patient" to refer
both to the subject of the intansitive verb, and to the direct object of a tansitive verb (both
marked by the absolute case). Thus, in the examples above, "boy" is the agent, and "ball"
is the patient. In practice, it is very difcult to escape using such common terms as
"subject" and "object", especially in unambiguous contexts, even if these terms do not
really ft Sumerian.
There are many ergative languages in the world, belonging to a number of different
language families: many languages in Australia, many American Indian languages, the
Caucasoid languages (for example, Georgian), Basque, to name a few. However, none of
what are sometimes refered to as the "major cultural languages" of Europe are ergative,
and so the concept is unfamiliar.
There are two other important points about ergativity. First, the defnition given above
describes what may be called "minimally" ergative languages. However, ergatvity can also
be refected in other pars of a language's grammatical system - it may affect verbal
agreement, cross-referencing of case-markers, coordination and subordination, etc. This
will be discussed in more detail later.
Second, there appear to be very few (if any) "pure" ergative languages. Most (perhaps
all) ergative languages are "split". I certain constuctions, the language behaves in an
ergative manner; i other constructions, the language behaves in an accusative manner. I
Sumerian, for instance, the perfect aspect functions i an ergative manner, while the
imperfect aspect functions in an accusative manner. That is, Sumerian is split along an
aspectual axis. There are other languages in the world which are split along exactly such an
axis, that is, the perfect aspect functions in an ergative manner, and the imperfect aspect
Classifcation 9
functions i an accusative manner. Also, the independent prnouns in Sumerian functon
basically on an accusative, not an ergative, basis. Languages of the world show a rather
bewildering variety and complexity in the ways that they are split.
In addition, there are languages which use an ergative - absolute differentaton to
mark semantic distinctions which are not easily made in the Semitic or Indo-European
languages. An oft-cited example is the sentence "We fell" in Bats, a member of the
Caucasoid language family, spoken in Georgia. If the act of falling is purely an accident,
outside of our contol, the subject of the sentence is in the absolute case. If we fell as a
result of our own action, the subject is in the ergative case. Other languages use an ergative
- absolute differentiation to mark other kinds of inforation, such as degrees of animacy.
Because there are very few (if any) pure ergative languages, it is perhaps best not to
think of "ergative - accusative" as a simple binary oppositon. C.T. van Aalderen has said
that "One suspects that the whole phenomenon is more a continuum than a set of
oppositions" ( 1 982:27). That is, some languages are closer to one "pole" than to the other.
Several recent linguists, for example, speak of "degrees of ergativity" in different
languages.
In the last twenty years or so, general linguists have shown a great deal of interest in
ergative languages; the bibliography of recent works is vast. In one of the more recent
articles, John Du Bois says:
Seemingly, ergatvity stands as a challenge to the view that all languages are
built on one universal archetype . . . . Why are there ergative languages in the
world? . . . Ergatvity . . . would seem somewhat pererse in splitting up an
apparently basic category like subject, assigning half its contents to a
contasting category like object. This perception of unnaturalness is of
course only an index of our failure to apprehend the actual basis of ergativity,
a diffculty which is simply reinforced by taditional grammatical
terminology ( 1 987 : 805-7).
It is only somewhat recently that the term ergatve has been systematically used for
Sumerian. Although some early researchers had intimations that this was how Sumerian
worked (even if all the details were unclear, as they still are), it is only in the last few years
that ergativity has been explicitly discussed in Sumerian. This means that in reading even
fairly recent Sumerological literature, such concepts and ters as "ergative", "agent",
"patient", etc., may not be used at all. The material might be discussed in what would now
be called an ergative model, without use of the term ergative, or in older works the material
might be presented in an accusative model. Moreover, not all scholars believe that
Sumerian functions on an ergative basis. Some Sumerologists believe that not enough
evidence has been presented to prove the case, and also believe that there are too many
"exceptions" to the model. Others disagree on the degree to which Sumerian can be said to
be split. Given the complexities of split ergativity in the languages of the world, it may be
that curent presentations of ergativity in Sumerian are too simplistic. "Full" proof can only
be forthcoming when there is more secure knowledge of Sumerian verbal morphology.
The frst person to apply the term ergative to Sumerian was apparently Viktor
Christian in 1 957, although he used the ter a little differently than it is usually understood.
10 Manu of Suera
Diakonoff ( 1965) sketched the system of ergativity in Sumerian and other Ancient Near
Easter languages, without explaining the details of morhology. The articles by Daniel
Foxvog ( 1 975) and Piot Michalowski ( 1980a) viewed Sumerian in an explicitly ergative
framework, while elucidating the verbal morhology. Van Aalderen (1982) has explored
some of the theoretical issues in more detail. The grammar by Marie-Louise Thomsen
( 1984) also follows a split-ergative model. A recent surey of ergativity in Sumerian is by
Gong Yushu ( 1987).
Agglutination
Sumerian is often described as an "agglutnatve" language. This term goes back to the
nineteenth century, when linguists attempted to classify the languages of the world into a
few basic types, based solely on typological (not genetic) criteria. For these linguists, the
three most common types of language could be classied as:
Isolatng
I isolating languages, virtually every morheme forms a separate "word". I
Chinese, for example, there are no tense-markers on verbs; such information is conveyed
by separate adverbs. There are also no plural-markers on nouns or verbs; this information
is conveyed by separate number-words.
Fusional
I fusional languages, such as Akkadian or Latin, grammatical morhemes are
expressed through endings on nouns or verbs, and several different morphemes tend to
"fuse" together. Latin amo, for example, means "I love". The /0/ ending on the verb
signals several things: the verb is frst person, singular, present tense, indicative mood,
active voice. However, none of the morphemes for person, number, tense, mood, or voice
can be segmented out -they are all fused into the ending /0/.
Agglutinative
I agglutinative languages, as in fusional languages, several grammatical morphemes
are combined into one word. However, the morphemes are distnct fom each other; they
do not fuse together. I an agglutinative language, stings of prefxes or sufxes tend to
occur; each afx is formally distinct, and expresses one morpheme. The parade example of
a language of this type is Turkish. In Turkish, the phrase "frm his houses" is expressed
as: evlerinden. Ev means "house", ler is the plural marker, in is the possessive pronoun
"his", and den is the postposition expressing the ablatve "from". I general, each afx
expresses one morheme; each morpheme is invariant: ler is the automatic plural marker
for all nouns; den means "fom" after any nominal phrase, etc. The morphemes are distinct,
not fused into each other.
Sumerian is similar to Turkish. The verbal phrase, for example, consists of a sting of
prefxes, followed by the verbal root, and then a smaller sting of sufxes. Each affx
expresses one morpheme, and each affx is (basically) invariant. Nominal phrases can be
very long, with a noun, modifying adjectives and appositves, genitive phrases, etc. , with a
Writing system 1 1
case-marker at the end of the entire nominal phrase.
The typological scheme presented here has ben somewhat simplifed. Moreover,
languages only tend to one category or the other; they are not "purely" isolatng, fusional,
or agglutinative. English, for example, is largely isolating, but it is also to some degree
fusional. It is occasionally agglutinative in its processes of word formation. In English
words such as "predictability" or "antidisestablishmentarianism", it is fairly easy to separate
several different morphemes, both as prefxes and as suffxes.
Most modern linguists who specialize in linguistic typology are not very interested in
this particular "morphological typology". They believe that such a scheme is not especially
useful, because it does not offer any interesting or helpful intuitions or generalizations
about language. The methodological underinning of this classifcation scheme has also
been attacked on severl grounds. For example, it was mentioned above that languages do
not usually fall neatly into one of these types. However, since the term agglutinative is still
used in Sumerological literature, especially in popular descriptions of the language, it is
useful to have some idea of what the term means.
The two terms ergative and agglutinative refer to different categories. The ergative -
accusative distinction depends on how the primary participants in a sentence are marked in
relation to each other. The isolating - fusional - agglutinative distinction refers to the
different ways that morphemes are combined into words. In theory, a language can be
either ergative or accusative, and also either isolating or fusional or agglutinative, although
not all of these possible categories seem to occur.
WITIG SYSTM
External chaacteristcs
In discussing any writing system, there are two factors to consider: the exteral
characteristics of the script, and the principles behind the script.
Because of the exteral shape of the signs in the Sumerian script, its writing system is
called "cuneiform". "Cuneus" is the Latin word for "wedge"; the term was coined because
of the most striking characteristic of the script -the fact that the signs are built up of stokes
looking like little wedges. (The term cuneiform was apparently frst used by one Thomas
Hyde in 1700. In his Historia religionis veterum Persarum, he refers to "dactuli pyra
midales seu cuneiformei".)
The cuneiform signs were inscribed by means of a stylus probably formed fom an
actual reed (such as still grows in modern-day Iraq), by impressing the stylus upon a tablet
of moist clay (or, occasionally, upon other surfaces). The stylus could also be made of
bone, metal, hardwood, or even other material.
The frst cuneiform texts discovered were all relatively late, fom a period when the
wedge-shaped characteristics of the script were most stiking. In the earliest phases of the
script, however, this wedge-shaped character is less pronounced; the script of most of the
Ur III inscriptions in this book does not look nearly as wedge-shaped as do later texts.
1
2
Mau of Sueria
The ten cuneifon refers solely to the exteral shape of the individual signs.
Cuneifon script was adopted and modifed by many peoples of the Ancient Near East; it
was used to write Akkadian, U garitc, Hurrian, Persian, etc. However, the fact that these
languages use signs with the same general exteral characteristics says nothing about their
possible genetc relationship. Sumerian, Akkadian, Hurrian, and Persian, for example,
belong to four entirely unrelated language families. Expressions such as "cuneifon
language" are occasionally encountered, but this is a rather imprecise way of refering to
one or several languages, which may or may not be related, which use a script with the
same exteral characteristics.
Ogia natue
The writing system used for English is an attempt to render speech as closely as
possible. Although English does suffer from numerus archaic spellings, and there are
certain features (such as upper and lower-case letters) which are found only in writing,
writing is basically an attempt to reproduce speech sounds. By contast, the Sumerian
writing system was never an exact, phonetic representation of speech; it was not "designed"
to reproduce spoken language as such. Rather, to some degree the writing system is only a
mnemonic device, to jog the memory of the writer and reader. The earliest uses of writing
were for administative texts, which were of a fonulaic nature, and whose contents were
familiar to the scribes. There was no need to write down what would be obvious to a
scribe who was a native speaker of Sumerian, and who was familiar with the material being
written. When such scribes "read" the texts, they knew how to supply the infonation not
indicated explicitly in the writing.
Thus, a certain amount of infonation in the spoken language was not expressed in the
writing. The further back in time one goes, the less the Sumerian writing system expresses
grammatical elements which are assumed to have been present in the spoken language. For
example, the basic graphic shape representing the rot for "to build" was originally a
picture of a wooden peg. I the earliest Sumerian, this one sign could be used for any
infected fon of the verb: any tense, mood, or person. Similarly, the expression for "on
that day" in Sumerian was: ud-bi- ("day-that-on"). But in the earliest Sumerian, only the
ud-sign was written; the reader inferred the rest.
As might be imagined, this lack of explicitess in the script can cause much touble in
interpreting Sumerian texts. Nor is this problem limited to the earliest Sumerian texts; in
late economic texts, for instance, it is often difcult to tell if something is being distibuted
"to" or "fom" somebody.
As time passed, the scribes wrote more and more down, that is, the writing became
more and more explicit. For example, ther is a Sumerian text known as the "Kesh Temple
Hymn", attested in several copies mostly frm the Old Babylonian period (dating to around
1 800 BC). In the 1960s, a version of the same text was found at Tell Abu Salabikh, datng
to about perhaps,2500 BC. Unfortunately, only a few lines of the Tell Abu Salabikh
version survive. But if one compares the Old Babylonian version with the Tell Abu
Salabikh version, it can be seen that although the text itself is relatively stable, the Old
Babylonian version indicates more verbal afxes than does the Tell Abu Salabikh version.
Writing system 1 3
This increase i n explicitness may be connected with the fact that Sumerian was gradually
dying out, and so scribes needed more help in their own understanding of texts.
Thus, a fundamental feature of the Sumerian writing system is its lack of explicitness.
It does not fully represent the spoken language. This has been summarized by Jacobsen:
"The history of Sumerian writing is one of progressively ever greater but never quite
attained adjustment to Sumerian speech" ( 1 957: 366 n. 1) . Similaly, Marvin Powell has
pointed out that "We fnd taces of its mnemonic chaacter enduring to the very end of the
Sumerian orthographic tadition" ( 1981 :421 ).
A further complicating problem is that the writing system is to some degree
morpheme-bound. There is indirect evidence to show that there were certain phonological
changes which took place in Sumerian, such as contraction, vowel deletion, etc., but these
changes ae masked by the script; the script often reproduces the basic morpheme, without
showing the changes which ae assumed to have taken place in the spoken language.
The view here presented, that the Sumerian writing system in origin and in practice is
basically mnemonic, has been especially expounded by Diakonoff ( 1976) and Stephen
Lieberman ( 1977).
Interal principles
The script used for writing Sumerian is a combination of "logographic" and "syllabic"
elements. Logographic means that a sign stands for a paticula word. For example, the
sign 4 stands for the word utu, "sun"; the sign stands for the word digi, "god". The
external shape of many of these signs is clealy pictographic in origin. Thus the sign for
"sun" was originally a picture of the sun rising over a mountain. The sign for "god" was
originally a picture of a sta. The original signifcance of many signs cannot yet be
determined.
The same sign can often have more than one logographic value. Thus, the same sign
can represent di, "god", or it can represent an, "sky". In general, it is only the context
which determines the meaning of the sign, and its correct reading.
Syllabic signs ae used to reproduce a sequence of phonetic elements. For example,
the sign is used to represent the syllable Iga/. This paricular syllable can form a
component of several different morphemes: it may be pat of the cohorative prefx on
verbs, or par of the ending of a genitive phrase on nouns, etc. The sign in these
contexts does not stand for any paricula word; rather, its purpose is to represent the
phonetic sequence I gl -I ai, which may form par of a number of different morphemes.
Syllabic signs can represent several different kinds of segments of consonants and
vowels. Some syllabic signs stand for single vowels, e.g. , and i. More common are
signs standing for the sequence consonant-vowel (ba, mu) or vowel-consonant (fli.
There ae some signs that stand for consonant-vowel-consonant, but these ae not common;
instead, the script uses a convention that represents ICVCI by CV-Vc. For example, the
segment Inirl is written by: ni-ir. A writing such as ni-ir does not imply a long vowel;
this is purely an orthographic convention, to reduce the number of CVC-signs which
would otherwise be necessay.
Many signs have more than one syllabic value. Many signs have both logographic
14 Mau of Suera
and syllabic values - sometmes more than one of each. The corect value of the sign can
usually only be derived from the context. Signs with more than one value are called
"polyvalent", or are said to have several "readings".
Thus, the Sumerian writing system is both logographic and syllabic. The syllabic
value of most signs derives fom a logographic value. For example, the sign in its
meaning as "sky" is pronounced lan/. This phonetic value was then generalized, so that
this sign can stand for the syllable lanl in other contexts.
I general, lexical morphemes are written logographically, and grammatical
morphemes are written syllabically, but this is not always the case. The system is
complicated by the fact that certain syllabic signs tend to be used for certain morphemes.
For example, there is a "conjugation-prefx" on the verb, pronounced Ibi!. There are
several different possible ways that this phonetic sequence could be represented in the
script. In practice, however, the scribes almost always used only one of these possibilities,
the sign. That is, certain morhemes tend to be indicated in only one way, and,
conversely, certain signs tend to be used only for certain morphemes.
I addition to logographic and syllabic signs, there are a few other elements present in
the script. One of these is "determinatives". Determinatives are signs which are used to
indicate the general semantic class to which a following (occasionally a preceding) noun
belongs. For example, almost all divine names are preceded by the sign ; this sign tells
the scribe that "what follows is a divine name". Most names of counties are followed by
the sign .this sign tells the scribe that "what precedes is the name of a country".
Determinatives were probably not spoken, even when Sumerian was read out loud. They
were only a feature of the written language.
I other contexts, the cuneiform signs which function as determinatives can also
function as logo graphic or syllabic elements. For example, the sign can represent digir,
"god"; the sign can represent ki, "county".
To sum up, Sumerian is mostly logographic, and only partially syllabic. Akkadian, on
the other hand, is mostly syllabic, and only partially logographic. Persian cuneiform is
almost entirely syllabic, and Ugaritc cuneiform is basically alphabetic. In practice, people
sometmes confuse the issue, and the term cuneiform is occasionally used to refer in general
to any logographic-syllabic system of writing, but this is wrong; there are many
logographic-syllabic scripts which have existed in the world, which are not cuneiform.
This has been a somewhat simplifed discussion of the Sumerian writing system.
There has been much recent discussion about the script, mostly hinging on theoretical
questions, such as the difference between pictographic and logographic, or the degree to
which the script is morpheme-bound.
Transliteraton
When citing Sumerian texts, or when discussing Sumerian grammar or vocabulary,
Sumerologists do not generally reproduce the original cuneiform signs. Rather, they cite
the word or passage in tansliteration into Latin characters. Transliteration is a sign-by-sign
image of the original written text. It is designed specifcally to refect the actual cuneiform
signs present. By looking at a tansliteration, one should be able to determine exactly
Writng system 15
which cuneiform signs occur in the original text (excluding palaeographic niceties).
Transliteraton seres several puroses. It is more convenient, quicker, and cheaper to
produce Latin characters than it is to produce cuneiform characters. Also, it provides an
approximate phonetic rendering of the signs occurring in the Sumerian. Since many
Sumerian signs have more than one reading, a scholar, by giving the text in tansliteration,
explicitly states his opinion about the reading of a particular cuneiform sign. For example,
the signcan be read Hkur (the name of a god), or im ("wind"), or ni ("self'
). Based
on his understanding of the text, a scholar decides the correct reading.
There are some complexities of tansliteration. It is possible for several different
cuneiform signs to have the same pronunciation. These signs must be differentiated in
tansliteration, so that the original cuneiform can be reconstucted fom the tansliteration.
For example, there are at least four different signs pronounced as /u/. If y were used as
the tansliteration for all four signs, it would not be possible to go backward from the
tansliteration: Given a tansliteration y, one could not tell which of the four possible signs
actually was written in the cuneiform. To obviate this problem, scholars have devised the
following system: The most common (or most important) sign with a particular value is
unmarked. The second most common (or most important) sign with this same value is
marked with an acute accent: g. The third most common (or most important) sign with this
same value is marked with a grave accent: !. The fourth, and higher, most common signs
with this same value are marked with subscripts: ., lS ' etc. This system is purely
arbitary; it provides a convenient means to differentiate between signs pronounced alike,
thus enabling us to reconstuct the cuneiform fom the tansliteration.
This use of the acute and grave accent-marks as "indices" has nothing to do with
pronunciation. They do not indicate anything about accent, nor do they indicate anything
about vocalic length, nor do they indicate anything about tone. They are used instead of a
possible Y
2
and Y
3
simply because it is easier to type accent marks (at least in Europe) than
it is to tur the typewriter carriage up to make a subscript.
These indices are based largely on frequency. However, these frequencies were
determined on the basis of Akkadian texts, not on the basis of Sumerian texts (for the
simple reason that Akkadian was "discovered" before Sumerian). This produces a certain
inconsistency. I Sumerian, for example, the bi-sign is much more common than the bi
sign. This inconsistency i s not really a problem; the only other alternative would have been
to devise a separate system for Sumerian, based on values and fequencies in Sumerian.
But this would have engendered so much confusion and complication that it is far easier to
work with the taditional system.
Confusion arises when indices are used on bisyllabic signs, that is, signs which
represent a segment of two syllables, such as /kala/ or /Urim/. I there is more than one
sign with the same bisyllabic reading, some scholars put the accent-marks on the frst
vowel, then continue onto the second syllable i there are several signs with the same
reading. Other scholars, however, begin with the last vowel, moving back to the frst.
Either system is prone to mechanical mistakes in printing, and the mere presence of the two
different systems can cause problems in deterining what the cuneiform sign actually was.
To mitigate against this difculty, some Sumerologists do not use acute or grave accent-
16 Mau ofSuera
maks on bisyllabic signs. Instead, they use a subscript
2
or subscript
3
when necessay.
For example, there are several signs with the value of /kala/. These are differentiated as:
kala, kala
2
kala
3
'
kala
4
' etc. This is the system followed here. Some recent publications,
including the Pennsylvania Sumerian Dictionar, use subscripts in place of accent-maks,
even on monosyllabic signs. Thus, instead of !, they use l
2
; instead of y, they use l
3
'
Determinatives ae a feature of the written language, and were probably not spoken.
To indicate that they were not pronounced, they are transliterated with superscript letters:
X
ki
,
tug
X, etc. For convenience sake, the determinative for god (the igir-sign) is tans
literated as a superscript
d
:
d
Bta. Because of the typographical diffculties of printing
superscripts, some publications instead print the determinatves on the same print-line,
connected by a period: X.ki; tug.x.
I tansl
i
teration, signs comprising one "word" ae linked by hyphens: kalam-ma,
il -ra-ni, etc. (Determinatives are an exception; no hyphens ae used.) As will be seen
below, it is not always easy to determine what constitutes a "word" in Sumerian.
Some Sumerologists use initial capital letters in their tansliteration of Sumerian proper
names; other Sumerologists do not. Those who do not use them, consider capital letters to
be a feature particula to the English writing system; since capital letters have no correlate in
the Sumerian writing system, they should not be used in transliteration. Other scholas feel
that since tansliteration is an artifcial device anyway, there is no ham in using capital
letters, if they help make the text cleaer to the reader. This second practice is followed
here.
Finally, it is necessay to say a few words about the typographic conventions used in
tansliterating Sumerian. Throughout this book, Sumerian is tansliterated by Roman
chaacters, underlined. The few Akkadian citations used here follow the same system.
However, it is occasionally inconvenient to use the same typographic conventions for two
different languages. To solve this problem, many publications cite Sumerian in Roman
chaacters, but widely-spaced. Thus, the word for "god" will be tansliterated as: digir.
This may seem like a convenient procedure to differentiate citations fom the two
languages, but it is prone to produce mechanical errors in printing.
It is fequently the case that it is not known how a particula Sumerian sign (or word)
is to be read. Some scholas elaborate the system just discussed, by presenting such
doubtful or unsure readings in caps. For example, the word for "interest-bearing loan" in
Sumerian is written: g. It is not sure how the frst sign is to be read. For this
reason, the word is often cited as: IAR-ra. Some scholas do, however, believe that they
now know how to read this word, and so nowadays one is likely to see the reading: urS
ra. That is, wide-spaced Roman is used for the "standard" tanslitertion of Sumerian, and
caps Roman is used for unsure readings. Not all Sumerologists follow this system,
however, and what is sure for one schola may be unsure for another schola.
Transcrpton
Transliteration is, by defnition, a refection of the written language, and so does not
necessarily reproduce the spoken language well (as we think we understand it). For this
reason, most Sumerologists use some form of transcription in their study of Sumerian.
Writingsystem 17
Transcription is not used as frequently as is tansliteration; it ocurs in discussions of
grammar, and appears in scholars' own notes.
Transcription attempts to reproduce Sumerian forms in their approximately correct
phonological and morphological shape, disregarding the omissions, conventions, and
idiosyncrasies of the written language. For example, signs appeaing as kalam-ma in
tansliteration, will appear as kalama in tanscription, since that is probably how the word
was actually pronounced.
There is no "offcial" or "standard" system of tanscription of Sumerian. It tends to be
somewhat personal and idiosyncratc, used by each Sumerologist to enable himself to
understand the language behind the written form. This situation contasts with that of
Akkadian, for example. In Akkadian there is a standard way of tansliterating texts, and
also a reasonably standard way of transcribing them. This can be done for Akkadian,
because scholars are generally confdent of their understanding of the rules of Akkadian
phonology and morphology; in general, tanscriptions of Akkadian done by different
scholars will be quite similar. In the case of Sumerian, there is much less confdence about
the language. Because the script does not always express all grammatical elements, the
morphology is not always sure. Moreover, there are several different analyses of the
phonetic structure of Sumerian.
The system of transcription used by most Sumerologists is not always transcription in
the precise sense of the term. For example, morpheme boundaries are ofen indicated.
Also, full forms of morhemes are often indicated, even when it is assumed that some
vocalic or consonantal segment probably dropped. Thus, it is actually a kind of mor
phological tanscription.
The system of transcription used in this book is based on the system of Jacobsen, and
is similar to what many Sumerologists use. It is a morphological transcription, in that it
separates morhemes from each other. In this system, morhemes are separated by
periods. Features which are assumed to have been present in the spoken language, but
which do not show up in the written language, are enclosed in parentheses. The different
indices which appear in tansliteration are ignored. Thus,

will be tanscribed as e, and
Urim
5
as Urim. Exceptions to this latter rule are sometimes made, parcularly for gram
matical morphemes which tend to be written in only one way. Thus, the "terminative"
case-ending is normally tanscribed by .e, because it is always written with the e-sign,
and never with the e-sign or the M-sign. Similarly, the "enclitic copula" is normally
transcribed as . am, since it is regularly written by the am-sign, and not by the am-sign or
the am-sign. (Details of these conventions will be discussed below.)
The difference between transliteration and tanscription should be kept in mind.
Transliteraton is essentially sign-by-sign, with the goal of representing the cuneiform signs
which were used in the original. Transcription is essentially word-by-word, with the goal
of approximating the correct phonological and morphological shape of a word. (In practice,
however, the terms transliteration and tanscription are occasionally used promiscuously.)
Transcription is important, because transliteration alone masks too many
morphological and phonological issues. Only a consistent tanscription can reveal a
thorough understanding of the language of the texts. Some of the simplest inscriptions, for
18 Mau of Suera
example, could be tanslated without knowing much Sumerian, simply fom a knowledge
of Akkadian and of simple vocabulary; a tanscription refects the stucture of the language
hidden beneath the written for.
At certain times in this book, the purely phonemic stucture of Sumerian will be
stessed, ignoring any morphological considerations. I that case, normal linguistic practice
will be followed, and the item will be put between slashes, e.g., Ikalama/.
Thus, our understanding of Sumerian may be refected in three different ways: a
tansliteration, refecting the written shape; a phonemic tanscription, refecting the
pronunciation; and a morphological tanscription, refecting our understanding of the
pronunciation and morhology.
PHONOLOGY
Poblems
It is not easy to reconstuct the phonological system of Sumerian, or the precise
pronunciation of any of its sounds. There ae two main reasons for this problem. Since
Sumerian is a language-isolate, there is no compaative evidence to provide help.
Moreover, most of the evidence for Sumerian phonology has been fltered through the
Akkadian phonological system; Sumerian phonology is seen through Akkadian eyes. For
instance, it is quite likely that the word for "son" in Sumerian was pronounced Idomu/,
with an initial lol-quality vowel. But Akkadian does not have an lol-quality vowel, and
hence no 101 -sign, and so this word is spelled out in syllabic Akkadian as: du-mu. If there
were only Akkadian evidence, it might never even be known that Sumerian had an 101-
quality vowel. Thus, the picture of Sumerian of the Ur III period (21 12-2004 BC) is
actually based on Akkadian of the Old Babylonian period ( 1 894- 1 595 BC), and later.
(Similaly, much knowledge of Sumerian grd a derives frm the interretations given to
it by Akkadian-speaking scribes and scholas; this topic i s discussed in Appendix 2. )
Likewise, very little is known about the historical development of Sumerian
phonology. Sumerian was spoken over a period of several centuries (and was used as a
written language for even more centuries). The phonological system of Sumerian at the
time of, say, Tell Abu Salabikh and that of the time of Ur II may have been signifcantly
different.
To some degree, more is known about the value and prnunciaton of Sumerian
grammatical morphemes, than about Sumerian lexical morphemes. This is because
grdatical morhemes ae mostly written syllabically, while lexical morphemes are
usually written logographically. Without the evidence of lexical lists (Appendix 2), it is
quite diffcult to fx the value of a logogram. For the same reason, it is occasionally
possible to see phonetic change through the course of Sumerian in grammatical
morphemes, but it is more diffcult to see such changes in lexical morphemes.
The upshot of this is that Sumerian prbably possessed sounds which Akkadian did
not, and which can only be deterined using a variety of indirect evidence. Because of the
diffculty of dealing with this indirect evidence, there have been several different
Phonology 19
reconstuctions of the Sumerian phonological system. These reconstuctions differ both in
the number of phonemes present in Sumerian, and in the value attibuted to certain
phonemes.
I practice, however, most Sumerologists do not t to exactly reprouce the sounds
of Sumerian. Rather, they use the standard values known from Akkadian. Thus, virtually
all tansliterations of Sumerian will use the value dumu for "son", even though this is one
of the clearest cases where an 101 -quality value can b postulated for Sumerian. Similarly,
it is sure that Sumerian had a velar I
l
I , which did not exist in Akkadian. The sign ,
for example, represents I
l
ul , the velar nasal followed by an lul -quality vowel; this is the
morpheme for the fst person singular possessive-sufx on nouns. But the normal value
of this sign in Akkadian is Imu/. Therefore, many Sumerologists tansliterate this sign as
mu, e.g., lugal-mu, "my king". Other scholars, however, tansliterate this sign as g!
I
O
e.g.,
lug-g
!
I
C
Still others, who wish to b more precise, in fact tansliterate this sign as
l
!lO'
or as some typographical equivalent, such as
1O
'
g
!lO'
etc.; for example, lug-g
!lC
This
means that tansliterations of Sumerian will differ somewhat fom scholar to scholar. The
tansliteration used here will refect the conventional method of tansliteration used by most
Sumerologists, even if this reconstuction is somewhat shaky and incomplete.
Vowels
Sumerian had at least the following vowels:
i u
e
a
The precise phonetic value of these vowels, particularly the I el, is unsure.
Many scholars also believe that Sumerian had an lol-quality vowel, but since no 101
existed in Akkadian (at least on the phonemic level), there is only indirect evidence to
reconstruct it. It is very difcult to determine whether any particular Sumerian word had an
101 -quality vowel or an lul -quality vowel; its existence has been established for only a few
cases. Under the assumption of the existence of this 101 -quality vowel, the vocalic system
of Sumerian is more symmetrical:
i u
e 0
a
Other Sumerologists have posited other vowels, such as both an open lel and a closed
le/. Others have posited the existence of nasalized vowels, but the exact number and
quality of these varies fom one scholar to another: /i/; le/; /i I and la/; /i I, lal and lel,
etc. Claude Boisson ( 1988) has investigated various reconstructions of the phonemic
system of Sumerian, in comparison with what is known about language in general. He
feels that i Sumerian possessed only four vowels, then the vowel normally reprsented as
20 Mau of Sueria
lel was more likely IEI than le/. He also feels that none of the systems of nasals which
have been posited for Sumerian is likely.
It is not sure if there was a phonemic distinction between short and long vowels; this
cannot be told fom the script. It has been postulated that there were no originally long
vowels in Sumerian, but that they did arise through vocalic contracton, in particular the
contaction of fnal root-vowels with initial vowels of suffxes.
As discussed above, in practcal terms most tansliterations of Sumerian usually only
refect the vowels known fom Akkadian; that is, the four vowels listed above.
Consonants
Most analyses of Sumerian would include the following consonants:
b p
d t
g k
Z s
b
m
n
I
1 r
(For ease in printing, the consonant indicated above as b is often simply tansliterated
as h without the "dish". Since Sumerian does not have a "simple" IhI, there is no
ambiguity in this usage. )
Virtually all Sumerologists accept the existence of the velar nasal III (although some
scholars prefer to speak of a palatal nasal, and others have seen more complex phonemes,
such as IIm/). When Sumerian words containing this phoneme are loaned into Akkadian,
it is usually (although not always) refected as !. For example, , "kind of priest"
(Lesson 21 ) appears in Akkadian as ang'.
Translitertions of this phoneme vary. I older works, and in many contemporary
works, it may simply appear as g. Some recent works use g, or some typographical
equivalent (g, etc.). It will be tansliterated here as g, in cases where it is assumed by most
Sumerologists to be present. With many words, however, it is not known whether a
phoneme is II/, Igl, or even 1nl or Iml, and so some variation in the tansliteration of
certain words appears. For example, the verb "to go" is understood by some
Sumerologists to be Iginl, but by others to be Iginl (or Igen/).
Many Sumerologists believe that Sumerian had a phoneme usually symbolized by
Idr/; its exact phonetic signifcance is unsure. Its existence has been proven in only a few
cases. Because of the diffculties of proving its existence in specifc words, it is usually not
indicated in tanscription; instead, in the standard sign-lists and in most tanscriptons it is
refected as g.
Several other consonants have been posited for Sumerian: Ihl, Iwl, Iy/; two (or
more) types of /11; two (or more) types of Ir/; a labiovelar Ik
w
/; a pre-nasalized labial stop
I
m
b/; etc. Since none of these sounds exists in Akkadian, the evidence for their existence
in Sumerian is indirect at best, and individual Sumerologists have their own preferences.
Phonology 21
Translitertons of Sumerian do not normally ty to reproduce these disputed phonemes.
As a typical example of a reconstucton of Sumerian phonology, it may be instctive
to present that postulated by Lieberman:
e i
a 0
u
b p m
d t n z
g k
g
z s
b
1 r f
I the tables above, certain consonants are indicated as differing only in voice: Ib/
Ip/; Idl - It/; etc. It is not in fact sure what differentiated such pairs; Lieberman explicitly
says that the distinction he marks as Ibl - Ipl was not one of voice. Some Sumerologists
have speculated that the difference was one of aspiration; this is not an uncommon view
today. Boisson, for example, says: "A correlation of aspiration seems to be the only
hypothesis with a high probability of success" ( 1988:25). Other Sumerologists have
speculated that the difference was one of glottalization.
There does not appear to have been a phonemic distinction between short and long
consonants; it is not in fact sure if long consonants occurred at all.
One of the thoriest questions in Sumerian involves the status of word-fnal and
syllable-fnal consonants. According to most Sumerologists, certain consonants, when in
word-fnal position, were not prnounced. For example, the root for "dais" is Ibarag/,
with a word-fnal Ig/. However, unless this Igl was followed by a vowel, it was not
pronounced: this word would have been pronounced as Ibara/.
The word-fnal consonant in a root is usually referred to by the German term
"Auslaut". Thus, it is said that the word for "dais" (pronounced Ibarag/) had a "g
Auslaut", or the word for "to live" (pronounced Itil/) had a "l-Auslaut".
The consonants which were regularly not pronounced in word-fnal position are called
"amissable" consonants. Those which were prnounced in word-fnal position are called
"non-amissable". (These terms are apparently peculiar to Sumerologists; they are not used
by generl linguists.)
Sumerologists differ among themselves about which consonants were not
pronounced. Some believe this affected all consonants, although perhaps not "to the same
degree". Others believe that it affected a smaller number of consonants (although no two
lists of such consonants seem to agree exactly). Also, it is not known if the amissable
consonants were not pronounced in word-fnal position only; most Sumerologists believe
that they were not pronounced in any syllable-fnal position. Ao Poebel, for example (the
real father of Sumerian grammar), states that "As a rule, an amissable consonant is dropped
whenever it stands at the end of a word or syllable" ( 1935: 147). Similary, Samuel Noah
Kramer says: "All fnal consonants in Sumerian are amissable. . . . The ter 'fnal
consonant' as here used includes the consonant at the end of a syllable as well as the one at
the end of a word" ( 1 936: 1 9).
22 ManualofSumerian
The existence of amissable consonants is certainly not impossible. There is a close
parallel in French: In spoken French, word-fnal consonants are not pronounced (under
certain conditions), although they still appear in the written form. A few Sumerologists,
however, are not convinced of the existence of amissable consonants. They interpret the
problem as bing orthographic in nature.
The reason this question is stll unresolved is because of ambiguities in the writing
system. At various points in this book, different pieces of evidence will be cited, some of
which seem to indicate that word-fnal consonants were pronounced, and some of which
seem to indicate that word-fnal consonants were not pronounced.
The existence of amissable consonants means that the cuneiform signs which represent
words with these amissable Auslauts have two values: a "long" value, which includes the
amissable Auslaut (e.g., kalag, Urim
S
' !l, and a "short" value, which does not (kala, Uri
S
'
il. With some signs, the long value and the short value have different indices, e.g., tU
[with diacritic] and ! [without diacritc]. This annoying situation is partially due to the fact
that indices were originally assigned on the basis of frequency in Akkadian, not Sumerian.
Some scholars transliterate Sumerian using basically only the long values; others
transliterate Sumerian using basically only the short values. Other scholars use both, the
choice being determined by syllabic conditions: the short form if word-fnal (or syllable
fnal), the long form if not. Others are less consistent, using a mixture of long and short
values. This latter practice is particularly true of less recent Sumerological literature, where
one fnds a mixture of transliteration principles, based primarily on customary readings of
the cuneiform signs. Such customary readings have arisen from the piece-meal growth in
understanding of Akkadian and Sumerian. For example, in 1940 Kramer published an
edition of the "Lamentation over the Destruction of Ur". This is a Sumerian poem, some
436 lines long, bemoaning the destruction of Ur at the end of the Ur III period; it was
written probably about a century afer its destuction. In his Introduction, Kramer says that
"The time is not yet ripe for a thorough and scientifc overhauling of the Sumerian system
of transliteration". Therefore, he "deems it best to follow the more or less established
usage". In this system,
In the case of signs representing roots that end in a consonant and may have
either the long or the short value (e.g., the signs for il(Q, "to call", du(g),
"good", etc. , which may be read either pad, dug, etc. or !l, dul
O
' etc.) the
transliteration uses the longer value in spite of the fact that the shorter is
scientifcally the more correct. Only in cases such as !(), "day", and a(g),
"heart", where the shorter value has become more or less standard, is that
value used in our transliteration, although the inconsistency in transliterating
the signs for il(g) and du(g) as pad and dug while giving those for !(g) and
a(g) as 1 and M is only too patent ( 1 940:6).
Kramer is obviously irked by this inconsistency, but feels that there is nothing he can
do about it. Although he wrote this passage almost fourty-fve years ago, some editors of
Sumerian texts still follow such customary usage. A compromise made by some
Sumerologists is to put the Auslaut within parentheses, e.g. , kala(g). However, if the short
and long forms have different indices, this can create confusion; some scholars transliterate
Phonology 23
as ti(1), others as ti(1). In this book, all word-fnal consonants have been consistently
tansliterated (and tanscribed) .
Other features
There were undoubtedly other features in the spoken language, which the writing
system only hints at. There is only marginal evidence, for example, to deterine word
stess, and it will not be dealt with here. Similary, there is only the most indirect evidence
for sentence-intonation.
Because of what is claimed to be a large number of homonyms in Sumerian, it has
several times been argued that Sumerian possessed phonemic tones. Diakonoff, for
example, says: "Sumerian was certainly a tonal language, or else the many homonyms
would have made spoken Sumerian quite unintelligible" ( 1983: 86). However, the evidence
is indirect and slight. In fact. many words which earlier Sumerologists believed to be
homonyms have been shown to contain different Auslauts, and so are not actually
homonyms.
PART lO: LESSONS I SUERN GRAMAR
Lesson 1
This frst text is a royal inscription of Ur-Nammu, the founder of the Ur III Dynasty
(ruled 2 1 12-2095 BC).
Sig-list ad voabular
I this and subsequent sign-lists, the signs are loosely organized according to function.
Determinatives are frst, followed by proper names, nouns, verbs, and syllabic signs.
Determinative preceding divine names (DNs). Transliterated by a superscript "d":
_ Determinative following geographical names (ONs). Transliterated by a superscript
"ki":
ki
} Nanna Nanna (DN, masc)
Nammu Nammu (DN, fer)
J
Ur_
d
Nammu Ur-Nammu (personal name [PN, masc)
4 _ Urims (Uris) Ur (ON)
r nin lady, mistess; "lord"
an heaven
}/lJ
ur man, warrior

lugal king

house
~
dt to build

na
`

ni
25
26 Lesson 1

ma

ke4
.
mu
" <
Especially for those who are primarily interested in West-Semitic, it is not always easy
to master cuneiform signs. In certain ways, however, it is easier to lear the signs of this
period than the signs of later periods. I later periods, the repertoire of possible sign
shapes becomes quite reduced, so that (superfcially) the signs of the Neo-Assyrian period,
for example, all look very similar. I the earlier periods, however, the signs are much more
distinctive, making them easier to lea.
However, one prblem in studying the signs of the early periods is the occasional
wide variation in exteral shape of the signs. For example, the sign for , "house", looks
rather different in Text 2 than it does in Text 1 . This variation is due to several factors:
nature of the writing surface, different scribal taditions at different scribal centers,
individual idiosyncrasies of handwriting, etc. The sign-lists and vocabularies attempt to
produce the basic or essential shape of each sign; the signs in the autographs are
reproduced exactly as published.
Notes
The Notes discuss some of the more important vocabulary items. Often, reference is
made to Akkadian words which were borrowed fom these Sumerian words. This practice
is open to methodological critcism, since Akkadian is not Sumerian, and there is no reason
to assume that Sumerian words always kept exactly the same meaning when placed into an
Akkadian context. But since normally much more is known about the Akkadian term than
about the Sumerian ter, it is still useful to examine the Akkadian equivalents.
Nanna The city-god of Ur. The large temple-complex at Ur discussed below was sacred
to him in particular. He was associated with the moon; nanna in fact means "moon".
I Akkadian, the word nannaru occurs, glossed by the CAD as: "luminary, light (as
poetic term, an epithet of the moon god and !Star)". This Akkadian word may be some
kind of blend or contamination between the Sumerian word nanna and the Akkadian rot
nawan.
Because of this Akkadian word, some earier Sumerlogists believed that the
Sumerian word had an Ir/-Auslaut, and so the name sometimes appears as Nannar.
However, there seems to be no inner-Sumerian evidence which would indicate such an
Auslaut.
The moon-god was also referred to as Zuen; this prblem will be further discussed in
Lesson 1 3.
The Mesopotamian scribes interpreted the cuneiform sign expressing his name as
consisting of two signs: the SeS-sign ( or } ) followed b_the ki-sign ,
Therfore, in older works the name is sometimes transliterated as: SeS-ki. Mor likely,
Lesson 1
2
7
however, the second element was originally the na-sign, functioning as a phonetic
complement of some kind.
Namu Not much is know about this goddess. However, she is described as "the mo
ther who gave birth to heaven and earh", and as "the primeval mother, who gave birth to all
the gods". It is thus possible that at one time she played a more important role in Sumerian
cosmogony.
The cuneiform sign which represents this name can also b read engur, which lexical
texts equate with the Akkadian apsu, the "watery deep" (see Lesson 14). The cuneiform
sign may be an abstact representation of this deep.
In some older Sumerological works, the two readings of this sign (Nammu and eng)
were not clearly differentiated. Therefore, the name of the founder of the Ur II Dynasty
sometimes appears as Ur-Engur, or Ur-Gur.
UrmS In English, "Ur". One of the more famous cities in southern Mesopotamia; the city
after which the Ur III period is named. The name of the moder site is al-Muqayyar.
The etymology of the name Urims is unknown. It is also not known how these two
particular cuneiform signs (presumably, the e-sign followed by the ab-sign) came to
represent the name.
Urims is the long value of the sign. The short value is variously tansliterated as Uri ,
Uri2, Uri3, or Uris. The oscillation in diacritics illustates the problem of diacritical marks
on bisyllabic signs.
The sign-lists in this book give the long value frst, followed by the short value.
Because both are encountered in Sumerological literature, it is necessary to know both
values, even though this seems like a totally unnecessary burden upon the student.
Sometimes, the name is written e-unug and not e-ab, in which case it should pr
perly be transliterated as Urim2.
The English equivalent, "Ur", derives from the Old Testament f'ur kasdlmf, "Ur of
the Chaldeans". Exactly how the Hebrew f'urf derives from the Sumerian fUrimf is
unsure.
nin In general, the Sumerian word for "lord" is en; the feminine equivalent, "lady", is nin.
(It is not impossible that the two words are etymologically related.) However, in older
Sumerian nin can also be used to refer to masculine entities. Perhaps at one time the term
was genderless. In the Ur III period, this usage can be considered an archaism.
ur The usual interpretation of this word is something like "man; warrior, hero". In
bilingual lexical texts, ur is glossed as amelu, "man", and as kalbu, "dog". ur with the
meaning "dog" is not uncommon in Sumerian texts. However, ur meaning "man" seems to
occur only in personal names; it does not have this meaning in actual texts (although the
compound ur-sag, "hero", presumably "man-head", is common).
One might guess that the ur-sign was originally a picture of a dog or some kind of
beast, but even the earliest attestations of the ur-sign do not look very animal-like.
lugal Etymologically, a compound of lu "man" and gal "great". This word is further
28 Lesson 1
discussed in Lesson 7.
According to 1.1. Gelb, "The Sumerian word has several meanings: a) a dwelling
house, even a room b) palace, temple c) family, clan d) household. The same meanings
occur also for the Akadian bItum" ( 1979b:2). In the sense of "temple", it can refer either
to one particular building, or to an entire temple complex consisting of several buildings.
I very recent secondary literature, it is occasionally tansliterated as: 'a.
dd Although du occasionally means "to build" de novo, it more often means "to rebuild".
It is especially fequent when describing the rbuilding of temples which had fallen into
disrepair. Usually, it is diffcult to tell in any particular text whether du means "to build" or
"to rebuild"; this can only be resolved by historical or archaeological data.
Gelb adds that "It is clear that when a ruler writes of having built a temple for a certain
divinity, he means not only that he erected a temple, but also that he provided it with all the
necessary means of social and economic support" ( 1979b: 3).
b
~

F
m

Lesson 1 29
Text 1

1
T
* "



"
---
--

30 Lesson 1
Notes: autographs
When obtainable, photographs of the texts used in the Lessons have been included.
This has not always been possible or desirable, and so most of the texts are presented as
"autographs". In Assyriological parlance, autograph refers to the hand-copy done by a
moder Assyriologist, to imitate the cuneiform. The quality of autographs can range frm
very accurate to very poor. To quote Lieberman,
It is, of course, patent that the "autographs" of all copyists are not equally
reliable. Their objectives, ranging fom an exact reproduction including
every scratch on the tablet to a highly abstact conventional representation of
the original (some Assyriologists are even known to have produced "copies"
fom their transliterated notes) as well as their individual skills and abilities
make the value of their copies diverge ( 1 977: 67).
It is only through long experience that one gets a feel for how accurate certain
Assyriologists are (or aren't) in their autographs.
-Writing practices
Both Sumerian and Akadian are written from left to right across the writing surface.
(The earliest Sumerian texts were inscribed in vertical columns, read from right to left.)
Most royal inscriptions are subdivided into "lines", marked by an actual line drawn or
impressed on the writing surface. The use of such lines in Sumerian (and Akkadian) is to
some extent dependent on the genre of text; royal inscriptions, for example, use them
regularly. Many literary texts use them, but just as many do not.
There is some oscillation in the use of the word line. This particular text was divided
by its scribe into seven units, but the ffth of these units actually contains two rows of text.
In order to be precise, some Sumerologists use the term "case" or "register" to describe the
units physically demarcated by the scribe, and the term "line" to describe the actual rows of
signs. Thus, in this text case 5 has two lines. Although this is a very handy distinction,
most scholars, will, in fact, simply use the term line to mean both line or case, especially in
unambiguous contexts; this is the _rcedure followed here.
In line 5, the deterinative begins the second line within the case. There are six
cuneiform signs in this particular expression. It would have been physically impossible to
put all these six signs on one line, so the scribe put them on two lines. If he had put the ki
sign with the Urim
s
-sign, there would have been too much empty space on the second line
of the case. By indenting the second line of the case, the signs representng the ON are
grouped in close proximity to each other.
Transliteration Transcription Translation
1 :
d
Nanna Nanna For Nanna,
2: nin-an-na nin.an.a(k) the "lord" of heaven,
3: nin--ni nin.ani. (r) his "lord" -
4: Ur_
d
Nammu Umammu Ur-Nammu,
ki
5: lugal-Urim
S
-ma-ke4
6: -f-ni
7: mu-na-du
Comenta
Lesson 1
lugal.Urim.ak.e
e.ani.0
mu.na.(n.)du.0
the king of Ur -
his temple
he built.
31
1. Nanna i s the name of a god; Nammu i s the name of a goddess. Sumerian has no gender
system; there are no special markers for either inherently masculine or inherently feminine
nouns. In most cases, one word may apply to either gender. For example, dim may mean
either "god" or "goddess". In other cases, the masculine and feminine seem to be formed
from different roots. In a few other cases, Sumerian adds the word for "female" (munus)
after a noun. For example, dumu can either mean "son" (masculine) or "child" (masculine
or feminine); dumu-munus is specifcally "female child", hence "daughter".
2. nin is used here to refer to the male god Nanna. For convenience sake, nin in such
contexts may be tanslated as "lord".
Sumerian has no defnite or indefnite article. For example, can mean "a house" or
"the house".
nin.an. a(k) forms a "genitive phrase". The formation of the genitive in Sumerian is
quite different fom the formations in Semitic or in Indo-European. In Sumerian, in a
genitive phrase consisting of two nouns, the "possessor" follows the "possessed". The
two nouns themselves are not formally marked, but the second noun is followed by the
"genitive marker" .a. For example, "the house of the king" is: e.1ugal.ak; "lady of
heaven" is: nin.an.ak. (Genitive phrases of more than two nouns will be discussed later.)
The form of the genitive marker is lakl following a consonant (in tanscripton, . ak)
and Ikl following a vowel (in tanscription, .k).
Ikl is one of the amissable consonants discussed under Phonology. As such, when in
word-fnal position, it does not show up in the writing system. As stated above, most
Sumerologists believe that the reason such consonants do not appear in writing, is because
they were not pronounced. A minority of scholars, however, believe that they were
pronounced, and their absence is purely an orthographic prblem. In the morphological
tanscription used here, the Ikl is tanscribed within parentheses: .a(k). This tanscription
shows that the Ikl does not appear in the script.
This genitive phrase is written nin-an-na, which is interpreted as: nin.an. a(k). One
might have expected a writing of the type *nin-an-f. However, Sumerian generally avoids
writing word-fnal (and to some degree, syllable-fnal) single vowels. Instead, the writing
system prefers to graphically reduplicate the consonant immediately preceding the word
fnal vowel. Thus, in this case, Sumerian writes the na-sign -graphically reduplicating the
preceding 1nl.
The principle of graphically reduplicating a preceding consonant is common
thrughout all periods of Sumerian. It is purely a property of the orthography; it does not
mean that Sumerian pronounced a double consonant her.
To summarize, nin-an-na represents the genitive phrase: nin.an.a(k). A genitive
phrase of two nouns is formed by adding the genitive marker after the second noun. The
32 Lesson 1
genitive marker is lal after a consonant, Ikl after a vowel. Ikl is one of the amissable
consonants, and hence does not appear in writing in word-fnal position. The lal of the
genitive marker is usually contained within a sign which reduplicates the consonant
immediatelyprecedingthe/a/.
3. nin-i-ni nin.ani.(r). Sumerian has a set of suffxes to indicate pronominal possession.
They are referred to as "possessive-sufxes" or "pronominal suffxes". .ani is the
possessive-sufx marking third person singular. The fons of the frst and second per
sons, and of all the plurals, are discussed later.
Since Sumerian has no gender system, .ani can mean either "his" or "her". However,
Sumerian does have remnants of what is usually referred to as a distinction in "animacy".
Human beings are "animate"; things and animals are "inanimate". In the case of the
possessive-suffx, .ani is only used to refer to animate antecedents; an entirely different
fon (.bi) is used to refer to inanimate antecedents (corresponding to English "its").
After a consonant, the suffx appears as: . ani. After a vowel, it appears both as: . ani,
and as: .ni. For example, "his house" can appear as both -i-ni and -ni; in the Ur II royal
inscriptions, the fuller spelling is much more common .
. ra is the case-marker for the dative case. Its fon is Iral following a consonant and
Irl following a vowel.
Case-endings in Sumerian work differently than they do in the Semitic or the Indo
European languages. I Sumerian, case-endings occur at the end of an entire nominal
phrase. A nominal phrase can vary in size. Minimally, it can consist of a single noun. It
can also consist of a noun with a possessive-suffx, or with an adjective, or with an
embedded genitive phrase, or even with a long series of appositives. I this partcular case,
the nominal phrase spans lines 1 to 3. It consists of: a divine name (Nanna); an appositive,
consisting of a genitive phrase (nin.an.a(k ; a second appositive, consisting of a noun with
a possessive-suffx (nin.ani). The dative case-marker .r comes at the end of this entire
phrase. This can be diagrammed as: [Nanna nin.an.a(k) nin.ani].r. This is, in general, the
way all case-markers work in Sumerian (and, even more generally, i agglutinative
languages).
The dative case is primarily used in Sumerian to express an indirect object; for
example, "He gave it to the king". It is also frequently used (as it is here) to exprss a
benefactive, that is, the person on whose behalf an action was perfoned. In such cases, it
can be tanslated by "for".
The case-marker Irl is not written here. Its presence in spoken Sumerian is shown by
the fact that it is actually written in other (mostly later) inscriptions. In these other
inscriptions, there are fons such as: dig-ra-ni-ir, "for his god" digir.ani.r (following the
nonal convention that CV-VC stands for ICVC/, that is, ni-ir Inir/). I the body of
texts in this book, .r frst appears in Text 14, an inscription of Amar-Sin, the grandson of
Ur-Nammu.
It is not known why the Irl is not written; this is discussed in Lesson 14. The
situation is different from that of the genitive marker. The Ikl of the genitive marker is an
amissable consonant, and so is regularly not written. But Irl is apparently a non-amissable
consonant, and does occasionally appear in the writing.
Lesson 1 33
4. ur followed by the name of a deity is a very common way to form personal names in
Sumerian, in all periods of the language. In Text 1 , Ur_
d
Nammu occurs; in Text 19a, Ur
d
Lamar. Such names are genitive phrases, meaning "man of DN" or "warrior of DN".
The name then is to be understood as: Ur.Nammu.(k), with the genitive marker taking the
form Ikl after a preceding vowel. However, there is some disagreement about the presence
of the genitive marker in proper names. Some Sumerologists believe that in proper names
the genitive marker was deleted. Thus, this particular name may have been pronounced as
lumammu/, and not as lumammuk/. Other Sumerologists however, do not believe this to
be so. The frst practice has been followed in this book, and thus this name has been
transcribed as Umammu, not as Ur.nammu.(k). In tanslation, the most common
Assyriological practice is to give the name as Ur-Nammu.
5. lgal-Urims
ki
-ma-ke4 lugal.Urim.ak.e, "king of Ur". Because the genitive marker
follows a consonant (here, Im/), its full form (/ak/, with initial la/) i s used.
When the genitive marker is directly followed by a vowel, the Ikl is pronounced, and
shows up in the writing (recall that in such phrases as nin.an.a(k), the Ikl is word-fnal,
and hence does not show up in the writing).
The .e is the marker of the ergative case, as discussed under Ergativity. As do all
case-markers, it comes at the end of the entire nominal phrse. The nominal phrase here
consists of a personal name, Ur-Nammu (a genitive phrase in origin), and an appositive
consisting of a genitive phrase, lugal.Urim.ak. This may be diagrammed as: [Umammu
lugal.Urim. ak].e.
The ergative case-marker .e marks what we would call the active subject of a transitive
verb, or, in more appropriate terminology, the agent. (Because of inconsistencies in
terminology, however, this .e is sometimes referred to as "agent", "agentive marker" or
"ending", "subject", "tansitive subject", "ergative marker", etc.)
The cuneiform signs do not refect well the morphology of Sumerian here. I
tansliteration, the signs are: lugal-Urims
ki
-ma-ke
4
' In morphological transcription, this is:
lugal.Urim. ak.e. The ra-sign reduplicates the fnal Iml ofUrims, and includes the lal of
the genitive marker. The ke4 -sign includes the Ikl of the genitive marker, and the I el of
the ergative case-marker. Thus, both the ra-sign and the ke4 -sign represent segments of
two different morphemes. This use of the ke4 -sign is very fequent; it is the sign normally
used for the combination of segments of the genitive marker and the ergative case-marker.
Not much is known about the syllabic stucture of spoken Sumerian, but it may have
been closer to the written form than to the morphological transcription. This line may have
been syllabifed something like: Ilu-ga-Iu-ri-ma-ke/. If so, the written form is actually
closer to the presumed syllabic stucture of Sumerian than it is to the morphemic structure
of Sumerian.
The use of hyphens in transliteration varies to some degree fom scholar to scholar.
All Sumerologists would use hyphens in the wo.d Urims
ki
-ma-ke4' Some would put a
hyphen between lugal and Urims: lugil -Uriml-ma-ke4
'
In this latter case, hyphens are
being used to link all the signs which form the entire nominal phrase. Others use hyphens
only between signs belonging to one word. It is not always easy, however, to defne
"word" in Sumerian.
34 Lesson 1
6. --ni e.ani.0, "his temple". As in line 3, .ani is the third person animate possessive
suffx. The antecedent is ambiguous; it could refer to Ur-Nammu, or it could refer to
Nanna. From other texts it is clear that .ani refers back to Nanna.
The .0 is the case-marker for the absolute case. This case indicates what we would
call the direct object of a transitive verb, or, more appropriately, the patient. There is,
however, not a great deal of consistency in nomenclature, and so such terms as
"accusative", "direct object marker", etc. , are commonly used.
The nominal phrase here is quite short, consisting of the noun , and the possessive
suffx . ani: [e.ani] .0.
7. mu-na-du mu.na.(n.)du.0, "he built". This line contains the verbal phrase. The verb
in Sumerian works rather differently than the verb in the Semitic or Indo-European
languages. A fnite verb form in Sumerian consists of a series of verbal prefxes, followed
by a verbal root, then followed by a smaller series of verbal sufxes. Certain of these
affxes are obligatory, while others are optional. Because of the general uncertainty of
Sumerian grammar, the precise number of prefxes occurring before the verbal root is
unsure. The view presented here might be called "minimalist". Alterative interpretations
will be discussed later.
The entire sequence of verbal prefxes occurring before the verbal root is usually
refered to as the "verbal chain". The frst prefx to appear in this chain is an optional
"modal-prefx" (also refered to as a "mood-marker"). Modal-prefxes are used for such
sentence types as cohortative, jussive, subjunctive, etc. A "normal" declarative sentence is
in the indicatve mood, which is unmarked. The verb in line 7 is indicative, and so there is
no modal-prefx.
The second position is occupied by the "conjugation-prefx". There are some half
dozen conjugation-prefxes. These prefxes are among the most mysterious features of
Sumerian; it is not known exactly what information these prefxes convey. This means
that it is not known, for example, what the difference in meaning is between a fnite verbal
form with the conjugation-prefx mu and one with the conjugation-prefx i. Such variation
occurs in the texts, but it is not known what this variation implies.
Needless to say, there are several theories about the function of the conjugation
prefxes. They may be connected with time: indicating whether events are near or far
(temporally, or even emotionally) relatve to the speaker. They may have to do with space:
indicating whether events are near or far (spatially, or even emotionally) relative to the
speaker. At times, they seem to corespond to a polite - familiar distincton.
It is probable that the conjugation-prefxes convey nuances which are not normally
conveyed in English. This means that even if it were understood what the conjugation
prefxes meant, it would not be possible to translate them readily into English, except by an
elaborate periphrasis. (Jacobsen, for example, believes that the conjugation-prefx mu is
used "To indicate 'closeness' to the speaker if by closeness we understand not only
closeness in space and time but also emotional closeness, empathy, involvement"
[ 1 965:
4
37] .)
In practce, Sumerologists ignore the conjugation-prefxes; they are not refected in
translation. Writing in 1972, Maurice Lambert said: "Today, the prefx does not exist for
Lesson 1 35
the tanslator of Sumerian, it is only an object of study for the grammarian" ( 1972-3:97).
The prblem of the conjugation-prefxes cannot be solved here. In subsequent texts,
the various conjugation-prefxes will be pointed out, and the possible kinds of informaton
which they may be conveying will be discussed.
Text 1 uses the conjugaton-prefx mu. This conjugaton-prefx is very common in the
Ur III ryal inscriptions. In fact, almost all past-tense verbs in main sentences in the Ur II
royal inscriptions use the conjugation-prefx mu.
The next set of prefxes are the (mostly) obligatory "dimensional-prefxes". There is
nothing comparable to these forms in Semitic or Indo-Eurpean. They "cross-reference"
(or "resume" or "register") the case relationships appearing in the various nominal phrases
in the sentence, with the exception of the agent and patient. In the verb in line 7, the
dimensional-prefx .na crss-references the dative case marked by .r in line 3.
Most earlier studies of Sumerian stated that the dimensional-prefxes were obligatory,
and that there was a one-to-one relationship between case relationships and dimensional
prefxes: every case relationship is resumed by its dimensional-prefx, and conversely
every occurence of a dimensional-prefx implies a coresponding case relationship
somewhere in the sentence. While this one-to-one corespondence may have been valid for
"pre-historic" Sumerian, in actual historic Sumerian the situation is not so neat. Gene
Gragg has made a detailed study of the dimensional-prefxes in the Old Babylonian literary
texts; he states that they "functon independently of concord to a much greater extent than
has been recognized by curent theories" ( 1973a: 10).
The dimensional-prefxes often seem unnecessary or redundant, because they do not
convey any new information; rther, they "merely" cross-reference the already-present case
relationships. However, all languages have a certain amount of built-in redundancy, to help
cope with the possiblities of information being garbled or lost. Many other languages
cross-reference case relationships, in various ways.
The nominal phrase in the dative is the only nominal phrase (except those indicatng
the agent and the patient) in the sentence, so only one dimensional-prefx occurs. If other
nominal phrases were present, they would also be resumed. Thus, it is possible for there to
be one, two, or three dimensional-prefxes in one verbal chain; that is, the dimensional
prefxes are cumulative. (The longest attested sequence appears to be four dimensional
prefxes in one verbal chain.) There is a hierarchical order to these prefxes; the dative, for
inst
ance, always comes frst. Not all such rules, however, are understood; in addition, there
are
certain morhophonemic changes which are not clear. These complicatons will be
discussed later.
Following the dimensional-prefxes comes a (prbably) obligatory prefx, the
"personal-affx" (there is no generally-accepted term). These forms have been much
discussed. They apparently cross-reference the agent and the patient, although this is not
completely certain.
In the case of a verb in the past tense, the personal-affx in this position cross
references the agent. Thus, in Text 1 , the personal-afx .(n) cross-references the agent
marked by the ergative case-marker of line 5.
The form of the third-person singular animate personal-affx is: .n. As will be seen
36 Lesson 1
later, the personal-affx has differnt forms for frst and second person, and also different
forms for inanimate agents.
This particular prefx always occupies the position closest to the verbal root.
However, this prefx frequently does not show up in the writing. The reason for its
absence is not as clear as that of, for example, the dative case-marker. The dative case
marker is not normally written in texts from the early stages of the Ur II dynasty (nor i
earlier texts), but it begins to show up fequently in texts from the time of Amar-Sin on.
Thus, scholars are reasonably confdent that the Irl of the dative case-marker is present,
even if not wrtten; its later appearance is the result of a change in orthographic practice.
The rules govering the presence and absence of the personal-affx 1nl are, however, not
so clear-cut; it is not often written even in later texts.
The presence or absence of .n cannot simply be correlated with a dimension of time.
In the Gudea texts, for example, forms both with and without .n occur, with no obvious
rules govering their distibution. And in later Sumerian, forms also occur both with and
without the . n. This means that rules cannot yet be determined for the presence or absence
of 1nl in the script, and it is not in fact sure at what level such rules would apply. The rules
may be purely orthographic; ther seem to be other cases in Sumerian where syllable-fal
nasals are not expressed in wrting. O, the rules may be phonological; the 1nl may have
dropped early, leaving a nasalized vowel, which could not adequately be represented in the
script. More probably, there may be a complex set of morphological and syntactical rules
govering deletion of 1nl; it has been posited, for example, that 1nl is only used (and so
only expressed in writng) to resolve possibly ambiguous cases.
Partially for convenience sake, I have assumed that the personal-afx .n is always
present, unless there is a specifc reason for its absence. Hence, it is tanscribed as: (n. ).
This presumed consistency must be taken with a grain of salt.
After all these obligatory and optiona prefxes, comes the verbal root, du in this
particular case. The root in Sumerian appears to be invariable. There is nothing like the
complicated infection of Semitic or Indo-European roots for person and number (the only
infection for person is in the personal-afx position, immediately before the verbal root; a
limited infection for number occurs in a set of persona-affxes after the verbal root).
There is no canonica shape of the root. Roots of the syllabic shape ev and eve are
perhaps the most common, but rots of other syllabic stuctures are fequent.
After the verbal root, there occur a number of optional afxes, not all of which are
well-understood. Some of these afxes are used to express modal and other nuances, such
as potentiality, irrealis, etc.
For a verb in the past tense, the most important affx which occurs in this position is
the personal-affx which cross-references the patient. The personal-affx which cross
references a third-person singular patient can be represented by zero, .0. Thus, the patient
in this sentence (e.ani. 0) is resumed by a .0 after the verbal root. This means that the
patient is ma
r
ked by .0, and that it is cross-referenced by .0. This may vaguely seem like
cheatng ("nothing resumed by nothing"), but there are theoretical justifcations for this
interpretaton.
Thus, the agent and the patient are resumed differently: The agent is resumed in the
Lesson 1 37
position immediately before the verbal root, and the patient is resumed in the position
im
mediately after the verbal root. The term personal-affx is used to refer to both affxes.
To sum up, the verbal phrase in Sumerian normally consists of: an optional modal
prefx (the indicative is unmarked); an obligatory conjugation-prefx, whose function is
unclear; one or more basically obligatory dimensional-prefxes, which cross-reference all
case relationships (except that of the agent and patient); an obligatory personal-affx, which
in the past tense cross-references the agent; the verbal root; an obligatory personal-afx,
which in the past tense cross-references the patient; other optional afxes.
This particular verbal form may be summarized as follows:
mu na . (n.) du . 0
( 1 ) (2) (3) (4) (5)
( 1 ) conjugation-prefx
(2) dimensional-prefx cross-referencing the dative
(3) personal-affx cross-referencing the agent
(4) verbal root
(5) personal-affx cross-referencing the patient.
The verb in line 7 was translated as past tense, without any discussion. Sumerian has
two sets of verbal forms. The difference in function between the two is somewhat unsure.
Some Sumerologists believe that the difference was one of tense (past - present-future);
others believe that it was a difference of aspect (perfect - imperfect); and others believe that
it was a difference of Aktionsart (punctual -durative, etc.). For convenience sake, they will
be referred to here as aspects.
Akkadian scribes gave names to these two aspects. One aspect they called bamm
("quick"), and the other they called maru ("fat"). There is some evidence that the Sumerian
word for bamm was Y g and the word for mari was niga; the original meaning of these
two words is not sure. The terms bam
n!
and mari are frequently used by modem Assy
riologists when referring to these verbal forms in Sumerian.
I the Ur III royal inscriptions, it does seem that basically the bamt.! is used for
actions which occurred in the past, and the mari is used for actions in the present and
future. That is, the two seem more tense-like than aspect-like. But this may be due to the
relative simplicity of these inscriptions.
The bamn! form is unmarked; it is the citation form (the form given, for example, in
the Vocabularies). As will be seen later, the mari is formed from the bamt! in several
different ways, and the systems for cross-referencing the agent and patient in the bamt! and
in the mari are quite different.
D
iscussion: structure
Having examined this inscription with a fne-tooth comb, let us now consider the
stucture of the inscription as a whole. If all appositional noun phrases are grouped with
their head nouns, and their functions are labeled, we see:
38
[Nanna, nin.an.a(k), nin.ani] . (r)
[Umammu, lugal. Urim.ak].e
[e. ani] .0
mu.na.(n.)du.0
Lesson 1
benefactive
agent
patient
verb
The dative marked in .r is resumed by the dimensional-prefx .na; the ergative marked
in .e is resumed by the personal-affx .n; the absolute marked in .0 is resumed by the
personal-affx .0. This is a rather aesthetically satisfying system; as will be seen later,
however, things often do not hang together so neatly.
Second, let us look at the word order:
BENFACTIVE - AGENT - PATIENT - VERB
( 1 ) (2) (3) (4)
This particular order is actually somewhat different from standard Sumerian syntax.
In more standard Sumerian, the word order is:
AGENT - PATINT - COMPLEMENTS - VERB
( 1) (2) (3) (
4
)
or: AGENT - COMPLEMENTS - PATIENT - VERB
( 1 ) (2) (3) (4)
The difference in word order between standard Sumerian prose and that of the royal
inscriptions is in the position of the benefactive. In royal inscriptions, the benefactive is
almost always fronted; this gives added emphasis to the deity on whose behalf some act is
being commemorated. In English, the difference might be reflected as "For Nanna, Ur
Nammu built his temple", instead of "Ur-Nammu built his temple for Nanna".
Hallo
'
s investigation of the stucture of the Ur III royal inscriptions showed that their
style is very formulaic. A typical inscription is composed of the following elements, almost
al ways in the same order:
( 1) A benefactive phrase, giving the name of the deity, with optional epithets;
(2) An agentive phrase, giving the name of the builder or donor, with optional
epithets;
(3) A patient phrase, describing the object built or donated;
(4) A verbal phrase, highly stylized and forulaic.
-Terminology
As does any discipline, Sumerology has engendered its own host of technical ters,
such as Auslaut, amissable, etc. Some of these terms are peculiar to Sumerologists; they
are not standard terms familiar to general linguists. Unfortunately, some of these terms are
used in ways which cause general linguists to take umbrage.
The term verbal chain is used here to refer to the series of prefxes which occur before
the verbal root. Other people use the term to include the entire verb: prefxes-root-suffxes.
Lesson 1 39
Similarly, the ten nominal chain is sometimes used to refer to a nominal phrase.
Sometimes both the verbal chain and the nominal chain are subsumed under the category
"Kettenbildung" .
The term conjugation-prefx, in particular, is misleading, because these elements have
nothing to do with conjugation, as this ten is usually understood. However, this is the
only term used by Sumerologists.
There is no standard ten to refer to what is called here the dimensional-prefx; the
most common ten is probably dimensional infx. This use of the ten infx, however, is
ofen irksome to general linguists, who use the ten to refer explicitly to an afx placed
within another morpheme; an example would be the It/ in the Akkadian Bt stem, or the
Arabic Eighth Form.
-ke4
The value of the _,sign as ke4 was deduced by Kramer in 1936. There is no native
grammatical tadition which gives this value; the Akkadian lexical lists give the values of
this sign as g and k4 (in addition to such values as kid, etc. ) Kramer reasoned that the
only way to make the Sumerian writing be consistent with our understanding of the
morphology of the Sumerian genitive was to posit a reading ke4
'
even if the lexical lists do
not give this value. Virtually all modern scholars have accepted his reasoning. (However,
even this seemingly well-established fact of Sumerian grammar has recently been
questioned, by Lieberman. He believes that the genitive marker was /ag/, not /ak/, but he
has not yet published his reasons for doubting the conventional interpretation.)
-Animacy
As was mentioned when discussing the possessive-suffx .ani, Sumerian has taces of
an animate - inanimate distinction. This distinction is also seen in the personal-afx of the
third-person bamt!-tansitive verb, where .n marks an animate agent, but .b marks an
inanimate agent (rather a rare occurrence). This animate - inanimate distinction does not
car y through all aspects of the grammar.
The tens animate and inanimate are those traditionally used by linguists, even if this
means that animals are called inanimate (Jacobsen prefers the tens "personal" and "non
personal"). In fables, however, animals are usually treated grammatically as animate.
-Conjugaton-prefxes
Lambert was quoted above, to the effect that the conjugation-prefxes are simply not
translated. This is because it is not known what infonation they convey, and the odds are
that their function has no easy equivalent in English. Edmond Sollberger has said:
Their true rOle is so distinctively Sumerian, they express ideas so alien to our
languages, that not only is there no consensus on the nature of their function,
but we simply ignore them without impairing, or so it seems to us, our
understanding of the text. There is no other translation for mu-gar and i-gar
than "(he) placed", although it must be pretty obvious that had there been no
difference there wouldn't have been two prefxes . . . . It is legitimate to posit
40 Lesson 1
that a certain verbal fonn implies that the acton is perfonned by the subject
wishing to indicate that his goal, though within his immediate perception,
remains without his actual sphere of physical contact; it is another thing to ty
and express that in one good English (or even Gennan) word ( 1 973: 1 60-61).
F.R. Kraus has criticized this view of Sollberger: "Sollberger's opinion, that Sumerian
texts can b understo without paying attention to the verbal prefx, is valid for a cerain
kind of text, but is cerainly not valid for legal documents" (1958: 83 n.47).
-Conjugation
The fonns of the bamm-tansitve verb in the singular are listed here. This and other
paradigms should b understood as refecting Ur III morphology, in Ur ITI orthography.
The model verb used is sar, "to write", with the conjugation-prefx mu.
frst person singular mu-sar mu.
0
. sar
second mu-sar mU.e. sar
third animate mu-sar mU.n. sar
inanimate mu-sar mu.b.sar
The fonn of the frst person is somewhat unsure. The fonn of the second person is
more sure, because the .e sometimes shows up in the script. Similarly, the fonns of the
third person are "sure", bcause of the occasional presence of .n and .b in the script.
I this section, the personal-afxes .n and .b have been discussed as markers for the
third person. Earlier, it was said that they cross-reference the agent. Stictly speaking, they
cross-reference a third person agent. A frst person agent ("I") is cross-referenced by .
0
,
and a second person agent ("you") is cross-referenced by .e. In other words, one can
understand the personal-affxes as crss-referencing the agent, or as marking the person of
the verb; in Sumerian, these are two different ways of describing the same thing.
-bamm and mai
The frst person to recognize that bamm and mai were used as native grammatical
tens was Heinrich Zimmer, in 1 885, although he did not know what they meant. Paul
Haupt was apparently the frst to give these words their etymologies as "quick" and "fat",
in 1932.
-Typology
The Intoduction discussed morphological typology, which is one attempt to bradly
categorize the languages of the world into a limited number of types. That particular
scheme of classifcation is somewhat out of favor, parially on theoretical grounds, and
partially oil the grounds that it does not offer more enlightening insights about language. A
more revealing scheme of linguistic typology is called "word order typology" (although
"constituent order typology" might be a more apt ten). This scheme examines the "basic"
order of the major constituents in a sentence. In English, for example, the most typical
order is: subject-verb-object. Hence, English is said to be a S-V -0 language. Sumerian,
Lesson 1 41
on the other hand, i s a S-O-V language.
S-O-V languages patter alike in several ways, not just in word order. For example,
very few S-O-V languages have prepositions. Instead, they use case-endings at the end of
nominal phrases, that is, "postpositions".
Also, in most S-O-V languages, adjectives follow their head noun, not precede. As
will be seen in Lesson 2, this is also how Sumerian works. Thus, in many ways (not in
all), Sumerian is a typical S-O-V language.
G. Haayer ( 1986) discusses some of the characteristcs of Sumerian in light of the
universal tendencies of language. He points out, for example, that "Most ergative
languages have SOY basic word order", and "The combination of ergativity and
postpositions in a single language points almost invariably to SOY basic word order"
( 1986: 80).
-Function of text
Let us now look at the function and Sitz im Leben of this particular text. Hallo has
divided the Ur III royal inscriptions into fve categories, based on typological criteria:
standard, building, votive, weight, and seal inscriptions. Text 1 is a building inscription;
examples will occur of all the other four types. Building inscriptions are defned by Hallo
as "monuments that became integral parts, whether functional or decorative, of the
buildings which they commemorated" ( 1962: 8).
The building inscriptions are further subdivided on the basis of the type of object they
were inscribed on: bricks (the most numerous of all royal inscriptions), foundaton
deposits, door sockets, and clay cones. Examples will be seen of each. Text 1 was
inscribed on a brick, forming an actual part of the masonry of a building. Building
inscriptions in general were not designed to be read by the builder's contemporaries; rather,
they were designed to be read by future rebuilders of the building, most likely future kings.
Ultimately, these buildings and their accompanying inscriptions can be thought of as
attempts by rulers to attain some form of immortality. (Text 16 is a door socket with two
inscriptions. One is of an early ruler of Ur [about 2400 BC]; the other is of a ruler of the
Ur il period. The door socket was evidently uncovered during rebuilding carried out in
the Ur III period, and was re-used.) As will be discussed in Lesson 2, ofen several copies
of the same inscription are found.
-History
Throughout Mesopotamian history, temples were built, repaired, modifed, or virtually
entirely rebuilt. During the Ur III period, there were many specifcally royal building
projects. Building inscriptions ofUr-Nammu have been found at Ur, Eridu, Larsa, Nippur,
and Uruk. He was responsible for building (and rebuilding) the large sacred area at Ur,
consisting of several stuctures. The most famous of these is its ziggurat, the best
preserved ziggurat in all of Mesopotamia: its base measures some 60 x 40 meters. It was
repaired by several later Mesopotamian rulers. (In Lesson 9, Woolley's reconstruction of
Nabonidus' rebuilding of this same ziggurat is pictured.)
The following drawing is Woolley's reconstruction of the ziggurat. The following
42 I0%On!
photo is of the remains now standing; the condition of these remains is partially a result of
modem reconstuction of the site.
Hc ZI CCURAT L UR,-NAMMU KLLKLL.

4 "
c88O0 ! 4
While the sacred area as a whole was dedicated to Nanna, he also had his own coun in
front of the ziggurat, and other buildings sacred to him. The entire sacred complex was
known as the E-ki-nu-Ml; the ziggurat was known as the E-temen-ni-guru3 (see Lesson
9). Both tenns are of uncertain etymology. The brick containing Text 1 foned pan of a
temple known as the -bur-sag ("mountain temple").
One of the more famous pieces of Ancient Near Easter art is known as the "Stela of
Ur-Nammu". It was found in a very fragmentary state in Ur, scattered throughout the
Nanna temple complex; it may have been destroyed during the Elamite sack of Ur in 2004
BC. It depicts a numbr of symbolic activities, mostly obscur to us, but apparently shows
Ur-Nammu himself carrying building tools (his name appears on a foating fragment of the
stela). This stela has been known since the 1920s, but restoration work is still on-going.
(A very interesting discussion is in Canby 1987.)
44 Lesson 1
In Lesson 7, a photograph is reprduced of a fgure which represents Ur-Nammu
himself (somewhat stylized) in his role as builder.
Discussing the function of the Mesopotamian ruler in this role as builder, Wolfgang
Heimpel says:
Te ruler in Mesopotamia, when building for the gods, manufactured the
frst brick himself, sprinkled the foundations with precious materials, laid the
foundation box, mixed some of the mortar, and led the celebrations of
dedication. Te best sources for these ceremonies are the building
inscriptions of Assyrian and Neo-Babylonian kings and the cylinders of
Gudea. Te latter contain the most detailed information which is couched in
poetic language and presents us with many diffculties of interpretation
(1987: 205).
-Literature
I addition to his well-documented role as builder, Ur-Nammu has become famous as
the promulgator of the world's frst attested law code, the "Code of Ur-Nammu". At least
three copies of parts of the text are known, but all are heavily damaged. The largest
fragment was found at Nippur. In 1 98 1 a fagment of the Code found at Sippar was
published. Basing himself on this new fragment, Kramer suggested that the "author" of the
Code was not Ur-Nammu, but rather his son Shulgi; this view has won general acceptance.
Ur-Nammu was also the subject of several literary works. These include "Te
Coronation of Ur-Nammu", a kind of self-laudatory hymn, and "The Death of Ur-Nammu
and his Descent to the Netherworld", in which his premature death on the battlefeld is
lamented.
-Proper names
Most recently, Miguel Civil ( 1985:27) transliterates the name of the founder of the Ur
III Dynasty as Ur-Namma, instead of the usual Ur-Nammu. He bases himself on
attestations of the name in syllabic orthography. He suggests that the original form of the
name was a theoretical IUr-Namnam/. Jacobsen also now reads the original form of the
divine name as Namma, but derives INammal frm Inin inim/, "lady female genital";
INammu/ is a later form ( 1987: 155 n.5).
As will be discussed in Appendix 2, a number of bilingual lexical lists have been
found at Ebla. Names of gods occur several times in these lists. For Nammu, the Eblaite
equivalent is given as: i-nu bf-mi-um. This is somewhat diffcult to understand. F.M.
Fales thinks that the Eblaite expression might mean "venemous tooth" ( 1 984: 176). It is
hard to square such a description or epithet with what is known about Nammu.
-Titulature
Many of the appositive phrases describing the king in these inscriptions are actually
titles, occurring in many inscriptions. (Although sometimes it is not possible to tell if an
adjectival phrase is a title or not.) Much work has been done in determining the origin of
certain titles, their relationship to parallel Akkadian titles, their falling into desuetude, etc.
Lesson 1 45
The
principal work on this topic is by Hallo: Early Mesopotamian Royal Titles: A Philo
mic and Historical Analysis (1957). This was followed by M.-J. Seux in 1 967, who
stud
ied in paticular the individual words occuring in Sumerian and Akkadian titles.
The title used in Text 1 , "King of Ur", was used by all fve kings of the Ur III
Dynasty.
Lesson 2
Text 2 is a second building inscription of Ur-Namu. It was inscribed on a brck,
fonning part of the Inanna temple (see Lesson 9).
Sign-list ad vocabular
AT
Inanna (DN, fer)
I. .J AJ
K
i-
m
-g
i
Sumer (ON)
i Ki-uri Akad (ON)
l nitab (nita) man, male
41
48
Lesson 2
I
kalag (kala) to be mighty

g
Notes
Inanna The daughter of Nanna. She was the Sumerian goddess of love and fertlity, of
the morning and evening star, and to some degree of war; she had other sides as well. She
may have absorbed some of the attibutes of originally independent deities. Later equated
with the Akkadian Ishtar, in some ways she was the most important goddess in the
Mesopotamian pantheon. Because of her rather fery temperament, and the manifold
aspects of her personality, she is perhaps the most interesting of all Mesopotamian deities.
She was worshipped in many cities, but especially in Uruk, where she was the tutelary
goddess. Her principal temple at Uruk was the Eanna -an-na e.an.a(k), "house of the
sky/heaven").
The reading of her name is much disputed. It is also tansliterated as: Inana, Innin,
and Ninni6' The original pictographic meaning of the cuneiform sign is also uncertain. Her
name is usually interpreted as: nin.an.a(k), "Lady of the sky/heaven". This is also how the
Akkadian scribes understood her name. Jacobsen believes that Inanna was originally the
"numen of the communal storehouse for dates". He thinks that the / an -component of her
name meant "date-clusters": "Her name . . . would appear to have meant originally 'the lady
of the date-clusters
'
" ( 1957: 1 08); later, her name was "re-interpreted" as "lady of the
sky/heaven".
Ki-en-gi This ON is always written syllabically. The etymology is unsure; this is dis
cussed below. The word ended in a /r/, not refected in the script. The Akkadian equi
valent of Kieng was Sumeru. This Akkadian word may be a dialectal pronunciation of the
word Kiengi(r). The English word "Sumer" is usually thought to derive frm the Ak
kadian form.
The frst appearance of Ki-en-gi is in an inscription of Enshakushanna of Uruk (who
ruled approximately 2432-2403 BC), who refers to himself as: en-Ki-en-gi l!gal-kalam
ma, "the lord of Sumer, the king of the land".
Ki-ur The etymology is unknown. It is not impossible that the ki-element was originally
a determinative.
nitab The basic meaning appears to be "male"; it can often be loosely tanslated as "man".
The Akkadian equivalent is zikaru, glossed by the CAD as: "1 . male (human and animal),
2. man, 3. ram".
kalag The Akkadian equivalent verb, dananu, is translated by the CAD as: "to become
strong". The verbal adjective, dannu, is translated as: "1 ) solid, strong, hard, heavy, thick,
massive, fortifed, steady, loud, 2) legitimate, binding, reliable, 3) stong, powerful, mighty,
great, 4) ferce, savage, diffcult, dangerous, serious, grave, obstinate, bad, tyrannical,
harsh, pressing, urgent, essential, imperative".
Lesson 2
Text Z
*

'| }

P
|

"
Notes
Some of the signs which occur both in Text 1 and in Text 2 differ slightly from ea
other. I Text 2, the -sign and the ke4-sign differ only in their length. In Text 1 they we
of the same length, but differed in the position and length of the verticals. Strictly speakir
the sign-shapes in Text 2 are more "correct".
I line 6, the word Ki-en-gi is divided into two lines within the one case.
5
0
1 :
2:
3:
4:
5:
Transliteration
d
lnanna
nin--ni
Ur_
d
Nammu
nitab-kalag-g
I I U
ki
.g- nm
5
-ma
Lesson 2
Transcription
Inanna
nin.ani. (r)
Urnammu
nitabkalaga
lugal.Urim.a(k)
6: lugal-Ki-en-gi-Ki-uri-ke4 lugal.KiengLKiuri.k.e
7: --ni
8: mu-na-du
Comentar
e.anL0
mu.na.(n.)du.0
Translation
For Inanna,
his lady
Ur-Nammu,
the mighty man,
the king of Ur,
the king of Sumer and Akkad
her temple -
he built.
2. nin--ni nin.anL(r), as in Text 1 . Here the nominal phrase expressing the benefactive
consists of the DN and an appositive, which itself consists of a noun with a possessive
suffx.
I this text, nin is used to refer to a goddess. This is the more normal practice; in Text
1 , nin refered to a god.
4. nitab-kalag-g nitab.kalaga. nitab is one of several Sumerian words meaning apprx
imately "man".
kalag-gi, representing Ikalaga/, is an adjective meaning "mighty". Many adjectives in
Sumerian end in la/, representing a morpheme .a. This . a has many uses: formation of
adjectves from verbal roots; nominalization of verbal phrases; marking of certain kinds of
subordinate clauses; etc. It is sometimes called a "nominalizer" or "nominalizing particle"
(although such terms do not cover all its uses). For convenience sake, the ter
"nominalizer" will be used here. In this particular case, the adjective Ikalagal is formed
from the verbal root Ikalag/, by the addition of the nominalizer I al. Since the nominalizer
.a is "built into" the adjective, it is not separated-out in transcription. That is, it is
transcribed as: kalaga, and not as: kalag.a. This is further discussed in Lesson 6.
Some Sumerologists prefer to say that Sumerian has no real (morphological) class of
adjectives, but instead has two kinds of "participles", one of which ends in the nominalizer
la/. For convenience sake, however, the traditional term adjective is retained here.
The two cuneiform signs of the adjective are here transliterated as kalag-gi. However,
the same two signs of this word are ofen transliterated as kala-gf!. If one looks in the stan
dard sign-lists for this particular sign ,_, it is given the values kal, kala, kalag, kalag,
and even kal.
Probably all Sumerologists would say that the word for "mighty" was formed from
two morphemes: the root Ikalagl with the addition of the nominalizer la/. They would
also say that the word was probably pronounced something like Ikalaga/. (There are late,
syllabic spellings of this word as kal-Ia-g, etc.) But exactly how do the two written signs
convey this information? There have been three approaches to the problem. One view is to
see the frst sign as representing the entire word Ikalaga/. In this case, the following -
Lesson 2 SI
sign would be a kind of "phonetic complement": It gives some exta information to the
reader, helping him to choose the correct reading of the previous sign. The word might
then be tansliterated as: kalag
ga
.
A second view attempts to make the signs approach the transcription. Since this word
is pronounced Ikalaga/, and since the I gal is expressed by the g-sign, this view says that
the frst sign must therefore be read Ikala/: kala-g. Thus, this view really derives the
tansliteration from the tanscription.
The third view says that the transliteration should not necessarily be expected to ft the
tanscription. Rather, there are certain general rules of Sumerian orthography which are
found in several different contexts. In this particular case, the transliteration kalag-g!
refects the orthographic rule that a consonant is graphically reduplicated before a word
fnal (occasionally syllable-fnal) vowel, particularly across a morpheme boundary. For
example, in Text 1 there occurred: nin-an-na, for nin.an.a(k).
The entire problem is not easy to resolve. Several obvious questions come to mind:
How can one know, for instance, that the sign _can, in fact, be read as kal, or kala, or
kalag, or kalag, or kalg? To what extent are readings "manufactured", to make the tans
literation more closely approximate the transcription? How valid is the general rule of
Sumerian orthography presented above?
In practice, inconsistencies arise in transliteration, because no matter which
transliteration system is followed, the meaning is normally clear. Whether these two signs
are understood as kalag
ga
, kalag-g or as kala-g, everyone would understand the pronun
ciation to be Ikalaga/, and the meaning to be "mighty". (Even here, however, some Sume
rologists would say that the original form */kalagal Ikalga/. It is true that similar cases
of vocalic loss are attested in Sumerian. However, the [late] syllabic writings of the type
kal-Ia-g! would seem to argue against such an interpretation in this particular case.)
Therefore, some Sumerologists prefer not to deal with these problems, unless they are
interested in the writing system per se.
This problem has been discussed at some length, because it is useful to be aware of the
theoretical principles which underpin our understanding of the writing system. This type of
knowledge is also essential if one is to understand borrowings of the Sumerian writng
system, such as, e.g., the writing system used for Eblaite. And, it is important to be
prepared for (and to understand the reasons for) the inconsistencies and variations in
transliteration which are encountered in Sumerological literature.
In general, adjectives in Sumerian follow the noun they modify.
5. ilgal-Urims
ki
-ma lugal.Urim.a(k), "king of Ur". Just as both kalag- and kala-g
are found in tansliterations of the same two signs, so also these signs are found tans
literated as Urims
ki
-ma and Uris
kC
ma.
6. lugal-Ki-en-gi-Ki-Uri-ke4 lugal.KiengLKiuri.k.e. , "king of Sumer and Akkad".
Sumerian has a conjunction meaning "and", linking nouns, but it is relatively
uncommon. Instead, Sumerian prefers to conjoin two nouns directly: an-ki "heaven and
earth".
The frst element of the genitive phrase is the single noun lugal. The second element is
formed by the two conjoined nouns, KiengLKiuri. The genitive marker . k follows the two
52 Lesson 2
elements. This can be diagrammed as: lugal. [Kiengi.Kiuri] .k. It is possible for either
element of a genitive phrase to be even more complex, consisting of a noun with a
possessive-suffx, an adjective, a relative clause, etc.
Lines 3-6 form a long nominal phrase, ending in the ergative case-marker .e. This
nominal phrase consists of: a personal name (line 3); an appositive consisting of a noun
and an adjective (line 4); a second appositive consisting of a genitive phrase (line 5); and a
third appositive, consisting of a more complex genitive phrase (line 6).
Discussion: stucture
It is instctive to compare the stucture of Text 1 and Text 2:
Text 1 :
[Nanna, nin.an.a(k), nin.ani] . (r)
[Urammu, lugal.Urim.ak] .e
[e.ani] . 0
mu.na.(n.)du.0
Text 2:
[Inanna, nin.ani] . (r)
[Urammu, nitag.kalaga,lugal.Urim.a(k),
I ugal.KiengLKi uri.k].e
[e.ani] . 0
mu.na.(n.)du.0
benefactive
agent
patient
verb
benefactive
agent
patient
verb
The order of the constituents is the same. As mentioned in Lesson 1 , the consttuent
order in these inscriptions is quite formulaic. The difference in the two inscriptions is in the
length of the various nominal phrases, and not in the basic stucture.
- Brick-stamps
The cuneiform signs in this text are much more "linear" than those of Text 1 . This is
because Text 1 was "handwritten" by a particular scribe. Text 2 was produced by a "brick
stamp". Brick-stamps were used to mass-produce copies of inscriptions. The writing on
them is done in reverse ("mirror-writing"), so that the impression comes out correctly. The
shape of the signs used tends to be linear, although occasionally they can approach the
shape of the handwritten signs. Several brick-stamps have been preserved, although
apparently none from the Ur III period. The following illustrations are of brick-stamps
from the Old Akkadian period:
! on 2 J
-Case relationships
.ra and the other case-markers in Sumerian are variously referred to as "cases", "case
markers", "case-endings", "postpositions", "postfixes", etc. Strictly speaking, these tenns
are not all synonymous, because they do not all refer to the same level of analysis.
The tenn "dative case", for example, refers purely to a grammatcal relationship. This
case can be used to indicate several different semantic relationships: indirect object,
benefactive, etc. "Dative case-marker" or "case-ending" refers to the specifc fonnal device
which signals this grammatical relationship, that is, the .ra. "Postposition" or "postf"
means that the case-marker occurs at the end of a nominal phrase. (This contasts with
English, for example, where "prepositions" occur in front of a nominal phrase.) Thus, in
Text i and 2, .ra can be described as a postpositive case-marker of the dative case, used to
express the bepefactive.
Although these tenns are distinct, in practice they are often used somewhat
indiscriminately. This is because it will normally be clear from the context which level of
analysis is being referred to. Similarly, the dimensional-prefixes are sometimes said to
54 Lesson 2
cross-reference the cases, and at other times are said to cross-reference the case-endings.
Strictly speaking, they cross-reference the case relationships which are marked by the case
endings. For ease of exposition, however, it is usually easier to present them as cross
referencing the case-endings themselves.
-Genitives and cases
The genitive does not behave like the (other) cases in Sumeran, and so it is
occasionally referred to as a "genitive marker", instead of as a case. First, a genitve phrase
can be embedded within a nominal phrase, which can then have its own case-marker. That
is, the genitve can be cumulative with respect to the (other) cases. For example, the
genitive can b directly followed by the ergative case-marker .e, as in Text 1 and Text 2.
The (other) cases, however, are not cumulative with respect to each other. If a nominal
phrase has the dative case-marker, for example, it is impossible for it to have any other
case-marker. Second, the genitive is not resumed by any dimensional-prefx. The dative,
for example, is resumed by the dimensional-prefx .na. However, the genitive is not
resumed.
The reason for the difference in behavior is because of the different r6le which the
genitive plays in a sentence. Genitives relate noun phrases to noun phrases. But the (other)
cases relate noun phrases to verb phrases. That is, genitives and cases perform two
different functions. However, "case" is the term most frequently encountered in
Sumerological literature.
Some scholars use the term "adnominal" case to refer to the genitive and to the
equitative (to be mentioned later). Both can be cumulative, and neither is resumed by any
dimensional-prefx.
The ergative and absolute cases pattern together, in that they are the only cases cross
referenced in the immediately pre- and post-verbal root slot. (In some ergative languages,
verbal cross-referencing only occurs with the agent and the patient, and not with any other
case relationship. )
The cases besides the ergative, absolute, genitive, and equitative are referred to as
"adverbial". They include the following; they will be studied in subsequent lessons:
dative; terminative; locative; locatve-terminative; comitative; ablative. Some scholars use
the term "oblique" instead of "adverbial"; others use the term "dimensional". The latter is
rather nice, since these cases are the only ones to be cross-referenced by the dimensional
prefxes.
To sum up, the Sumerian cases may be categorized as:
prmay
adverbial
(ergative; absolute)
(dative; terminative; locative; locative-terminative; comitative;
ablative)
adnominal (genitive; equitative)
-Earlier views of genitive
It was Poebel who defnitively established the form and function of the Sumeran
genitive ( 1935). Earlier views were quite different. For example, Franois Thureau-
Lesson 2
55
Dangin saw the genitive in Sumerian as being formed in two different ways: either by
simple "juxtaposition" of two nouns (lw-uru, "king-city" "king of the city"), or by an
ending .a of a "general indirct case". He thought that the Ikl which appears when a vowel
follows the genitive marker was "inorganic"; it was a "hiatus-breaker" to avoid a sequence
of two vowels. Poebel effectively destroyed Thureau-Dangin's views, but taces of the
latter are still encountered in some works. Poebel
'
s work was further elaborated by
Jacobsen (1973).
-Typology
Scholars have pointed out previously that the genitive in Sumerian behaves differently
than the (other) cases. This is typical of agglutinative languages, where the genitive will b
cumulative with respect to cases.
In most S-O-V languages, genitive constructions are expressed by the sequence
"possessor-possessed" (regardless of the exact morphological devices used). Sumerian,
then, would seem to be atypical, in that the sequence is "possessed-possessor". It will be
seen later that Sumerian also possesses a genitive construction of the type possessor
possessed, but this construction is not as common as the possessed-possessor one.
In Sumerian, modifers of nouns typically follow their head noun. In this lesson an
adjective follows its head noun, and in Lesson 7 a relative clause follows its head noun. It
is a general characteristic of S-O-V languages for modifers to follow their head nouns.
-Proper names
Many different etymologies of Ki-en-gi have been proposed, and just as many ex
planations for the derivation of Sumeru from Ki-en-gi. The sheer variety of such expla
nations shows how unsure such attempts are. Some of those proposed by more prominent
Sumerologists incl ude:
Aro Poebel: Ki-en-giCI) is a dialectal form of kalam "land".
Anton Deimel: Ki-en-giCI) ki.gir "land of the foot", i.e. "stopping place".
Thorkild Jacobsen: Ki-en-giW ki-Nigir; Nigir ) Nibir ) Nibur ) Nibru, "Nippur".
That is, the term "Nigir" (whatever this may have meant originally) was at frst applied
only to the city of Nippur. Later, in the form "ki-Nigir" ("place of Nigir"), it became
generalized to the whole land of Sumer.
Edmond Sollberger: Sumeru is the Emesal form of Ki-en-gi(r), whatever the ety
mology of the latter might be.
Edmund Gordon: Ki-en-gi ki.gir1
5
'
"noble place".
Many other dubious etymologies have been proposed. They illustrate the fact that
there is really very little evidence to make a positive decision; the data can be made to ft
many different interpretations.
56 Lesson 2
-Titulature
nitab-kalag-g! is a very old title, attested even with rulers preceding the Akkad dynas
ty. It was also used by Utu-hengal of Uruk. It is difcult to say exactly what an expres
sion like "stong man" or "mighty man" means; Hallo says: "'strong man' (that is, we
might almost say, independent ruler)" (1966: 1 38).
Ur-Nammu was the frst Mesopotamian ruler to use the title l!gal Ki-en-gi Ki-uri. It
was used by his son Shulgi, but not (apparently) by the other rulers of the Ur II Dynasty.
It was used sporadically by later rulers (in both a Sumerian and an Akkadian form), right
down to the Persian period, especially by conquerors of Babylonia (such as Cyrus).
Ur-Nammu frst assumed this title about the fourth year of his rule. I the early years
of his reign, the extent of his control was too limited, and his hold too weak, to perit use
of such a grandiose title.
Lesson 3
This is another building inscription of Ur-Nammu. Like many royal inscriptions. it
exists in several copies. To illustate the range of vaiation in the exteral shape of the
signs, two different copies of this one inscription are reproduced. Text 3a was prduced by
a
brick-stamp, as was Text 2. Text 3b was hand-written, as was Text 1. Both copies a
fom UT.
This brick is on display in the Britsh Museum. The display stand was obviously
made many years ago. It refers to Ur-Nammu as "Ur-Gur", it refers to Nanna as "Nannar",
and it dates the brick to "about B.C. 2500" - about four centuries earlier than today's
chronology.
"1
58 Lesson 3
Sig-list ad voabuar
bad city wall, rampart, fortifcation
Notes
bAd The PSD tanslates bad as: "wall", "fortifcation". Its normal Ak adian equivalent is
duru, glossed by the CAD as: "1 . city wall, fortifcation wall, 2. inner city wall, 3.
foress, 4. enclosure of a house".
Lesson 3
59
Text 3a
# N" T
H||


0 #
60 Lesson 3
Text 3b

n
p
"
t
^

I

t
Notes
Both inscriptions are inscribed on bricks, yet the signs i Text 3a are much more linear
than the signs in Text 3b. Because of the clearly wedge-shaped form of the signs, Text 3b
looks more like what we are accustomed to think of as cuneiform.
Differences between Text 3a and Text 3b also occur in lines 4 and 7. The various
signs which make up the geographical names are differently divided between the two lines
in each case.
1 :
2:
3:
4:
5:
6:
7:
8:
Transliteration
d
Nanna
lugal--ni
Ur_
d
Nammu
lugal-Urims
ki
-ma-ke4
--ni
mu-na-du
b' d U
' ki
-- nm
S
-ma
mu-na-du
Commenta
Lesson 3 61
Transcription Translation
Nanna For Nanna,
lugal.ani.(r) his king -
Umammu Ur-Nammu,
lugal.Urim.ak.e the king of Ur -
e.ani.0 his temple -
mu.na.(n.)du.0 he built;
bad.Urim.a(k).0 the city wall ofUr -
mu.na.(n.)du.0 he built.
1 . The frst six lines are essentially the same as those of Text 1 and Text 2. In line 2,
however, the ten lugal is used to refer to Nanna, instead of the nin of Text 1 .
7. The next two lines fon a new sentence, "he built the city wall of Ur". Line 7 i s the
direct object (patient), consisting of a simple genitive phrase.
8. The verbal phrase is exactly as in the previous inscriptions, and as in line 6.
The verb fon in line 8 contains a dimensional-prefx .na, and a personal-afx (n.).
However, there is no expressed datival phrase, nor agent, in lines 7 or 8. Rather, the logical
benefactive to which the .na refers is Nanna of line 1 , and the logical agent to which the (.n)
refers is Ur_
d
Nammu of line 3.
The syntax of Sumerian beyond the level of the simple sentence has yet to be
adequately studied. However, it often happens that a verbal chain will contain elements
which cross-reference nominal phrases occurring in a previous sentence.
Discussion: structure
The basic structure of this text is:
[Nanna, lugal.ani] . (r)
[Umaru, lugal.Urim.ak] .e
[e.ani] . 0
mu.na.(n.)du.0
[bad.Urim.a(k)]. 0
mu.na.(n.)du.0
-Sign fonation
benefactive
agent
patient
verb
patient
verb
Some cuneifon signs are, in origin, combinations of two different signs, one of
which is pictographic, and one of which is a phonetic indicator of some kind. For example,
the bad-sign is a picture of a city wall, with an inscribed bad-sign ( ); the function of the
inscribed bad-sig'n is to aid the reader in the correct pronunciation (another example occurs
in Lesson 15).
In Lesson 1, it was mentioned that the second element of the Nanna-sign , ,,which
62 Lesson 3
was interpreted by later Mesopotamian scribes as a ki-sign, may originally have been a na
sign, functioning as some kind of phonetic indicator.
-Co-ordination
In general, independent sentences in Sumerian are coordinated without any
conjunction. I Text 3, no conjunction appears in line 7. In Lesson 16 there occurs the use
of a conjunction g, borrowed from Akkadian.
- History
Most building inscriptions refer to only one undertaking (for example, the building of
a single temple), but it is not uncommon to fnd such inscriptions referring to two closely
related activities.
The wall referred to in this inscription was undoubtedly the wall which surrounded the
city of Ur. Woolley describes it as follows:
The walled city was in shape an irregular oval, measuring about 1 1 30 yads
in length by 750 yards in width, and was surrounded by a wall and rampart.
The rampart was of mud-brick with a steeply sloping outer face . . . . Along the
top of this ran the wall proper, built of burnt bricks . . . . Of Ur-Nammu
'
s wall
not a trace remained . . . just because the defences of Ur had been so stong
the victorious enemy [that is, the later Elamites who sacked the city] had
dismantled them systematically, leaving not one brick upon another
(1982: 1 37-8).
Although Ur-Nammu
'
s wall may have originally enclosed the entire city ofUr, the city
rapidly expanded beyond these walls. The original walled city may have comprised no
more than one-fourth or one-ffh of the city in the Isin-Iasa or Old Babylonian periods.
The situation must have been simila to that of many Middle Easter cities today, with a
core consisting of an "Old City" (often referred to as a "madina" in Arabic or English), but
with much extension beyond it.
Lesson 3 63
Text 3c
supplemen tary
The supplementary texts included here and in subsequent lessons are meant for prac
tice and review. They will nonally contain no new vocabulary or grammar; any new
features will be explained.
This particular text is another brick.
r
'f
r
f

Lesson 4
This inscription was engraved upon a stone bowl. No photograph is available.
Sign-list ad vocabular
_ a- Nin-gal Ningal (DN, fer)
MT>-
tU () to live
T . . . J
. . . I to dedicate a votive object
Notes
Nin-gal Ningal was the wife of Nanna, and the mother of Inanna. Being the wife of
Nanna, she was especially worshipped in Ur. Her name means "great lady".
65
66 Lesson 4
Text 4

"

=F

=
T

1 J
1 :
2:
3:
4:
5:
Transliteration
d
Nin-gal
nin--ni
Ur_
d
Nammu
nitab-kalag-g
I I U
ki
.g- nm
S
-ma
Lesson 4
Transcription Translation
Ningal For Ningal,
nin.ani. (r) his lady -
Urammu Ur-Nammu,
nitab

kalaga the mighty man,


lugal.Urim.a(k) the king of Ur,
6:
lugl-Ki-en-gi-Ki-uri-ke4 lugal.Kiengi.Kiuri.k.e the king of Sumer and Akkad -
7: nam-tU-Ia-ni-e
8:
-mu-na-ru
Commenta
nam. til.anUe
a.mu.na.(n. )ru
for the sake of his life -
he dedicated a votive offering.
67
7. Sumerian does not have many processes of word formation. However, it does have a
formative element nam. Prefxed to verbal and nominal roots, it prduces what are called
"abstact nouns". For example, l!al is "king", nam-l!gal is "kingship"; tU is "to live",
nam-tU is "life".
M is a case-marker not seen up till now, marking the "terminative" or "directive" case.
It does not have any one exact tanslation into English, although it generally indicates
"direction towards" or "action towards". Here, the meaning is something like "for the sake
of', or "on bhalf of'. This phrase means something like "for his life", that is, "so that the
king will live a long time".
The exact phonemic value of this case-marker is not actually sure. Since the dative
case-marker Iral ) Irl after a vowel, one might expect Iel ) IU after a vowel. Such a
reduction does occasionally happen, although not apparently in this particular expression
(the sign r does not appear to have any reading in liU). The original value of this
morpheme may have been, in fact, lee/. Conventional Sumerological practice is to simply
tansliterate it as e. It will be further discussed below.
The two signs transliterated here as: tU-la are also found transliterated as: !-la. This
is the same problem seen earlier with kalag-g - kala-g, and Urims
kC
ma - Uris
ki
-ma. In
this particular case, there is the added irritation that the sign l>Tis read ! (with no dia
critic) and til (with diacritic) .
8. . . . f "to dedicate a votive offering". This i s a "compound verb", a type of word
formation very common in Sumerian. A compound verb is made up of two elements. The
frst element of the compound verb immediately precedes a complete verbal form, including
its prefx chain. The second element functions as the regular Sumerian verbal root. In this
case, the frst element is ; then comes a regular verbal form, consisting of a prefx chain
and the verbal root ru; ru is the second element of the compound verb.
The original function of the two elements of a compound verb is often not clear. In
many cases, the frst element is (historically, at least) the patient of the verbal root which
forms the second element. For example, g! . . . de means "to speak". de means "to pour out",
and g! means "voice". Historically, then, this particular compound verb means "to pour
68
Lesson 4
out the voice"; that is, g! is historically the patient of de. In such cases, the frst or second
element may appear elsewhere as a free morpheme, functioning in other contexts like any
other noun or verb.
In other instances, the nominal element of the compound verb is in one of the adverbial
cases. For example, "to found" or "to establish" a temple or building of some kind is:
ki. . . gar. g means "to place"; ki means "ground". Here, ki is in the "locative" case (ki.a,
Lesson 6). The historical meaning was "to place (something) on the ground".
I some cases, the relationship between the nominal and the verbal components is not
sure. In this paricular case, for example, probably means "to send" (judging frm its
use in other contexts as a "simple", that is, non-compound verb), but it is not known what
the element means here.
Synchrnically, it is not easy to defne the term "compound verb". Certain verbs, for
instance, are almost always used with certain patients (cf. g! . . . de above). Are these
compound verbs or not? It i

diffcult to say; it is not easy to produce a rigorous defnition
of compound verbs. Compounds are basically identifed on semantic criteria. If the
meaning of the compound is more than the sum of its parts - that is, if it is loosely an idiom
- then it is felt as a compound.
I the sign-lists and vocabularies in this book, compound verbs are indicated by the
use of three periods: . . . ru.
In line 7, the terminative case-marker e occurs. According to what was said earlier
about the dimensional-prefxes, one might expect this e to be crss-referenced by a
dimensional-prefx. The dimensional-prefx which cross-references the terminative e is i;
it follows the datival dimensional-prefx in the prefx chain. Therefore, one might have ex
pected to fnd a form such as: -mu-na-i-(!)-ru.
This verbal form illustrates the basic problem of the dimensional-prefxes. Although
theoretically there is a one-to-one correspondence between case relationships and
dimensional-prefxes, in practice it isn
'
t so. Case relatonships are found that are not appa
rently cross-referenced by a dimensional-prefx, and conversely dimensional-prefxes are
found even when no case relationship is apparently present.
This is probably more than just an orthographic prblem. That is, it is not simply the
case that the dimensional-prefx is "there", but not written. In all the Ur III votive
inscriptions, the e in expressions of the type nam.til.anUe seems never to be resumed.
The presence or absence of dimensional-prefxes probably depends on semantic
factors at the level of the sentence, and at the level of the discourse, that is, beyond the level
of one single sentence. Certain nominal phrases are less closely bound to the sentence or to
the discourse than others. For example, nam.til. anUe is only loosely bound to the
sentence; it could be omitted, without any great loss of information. Such loosely-bound
phrases (almost idiomatic or formulaic in character) may perhaps not need to be resumed,
while such important constituents as the benefactive phrase would need to be resumed. Not
a great deal is known about the structure of Sumerian at this discourse level.
To sum up, although the problem is still open to discussion, it is more likely that it is a
semantic-syntactic problem, and not an orthographic problem. Therefore, no dimensional
prefx i is indicated in transcription.
Lesson 4
Discussion: stucture
The stucture of this text is:
[Ningal, nin.ani]. (r)
[Urammu, nitab.kalaga, lugal. Urim.a(k),
lugal.KiengLKiuri.k].e
[nam.til.ani].M
a.mu.na.(n.)ru
-Abstacts and concretes
benefactive
agent
purpose
verb
69
nam is the regular element used to form abstract nouns in Sumerian. nig is regularly
used to form concrete nouns from verbal roots; this is discussed in Lesson 22.
- Phonology of -M
The reason it is diffcult to determine the phonetic shape of e is because it seems to
appear in different forms even under identical conditions. For example, the writing u-zu
u ("towards your hand") would justify an interpretation as: u.zu.e, with the loss of the
fnal vowel, presumably conditioned by the presence of the lu/-vowel before the IU.
However, the form gir-zu-e ("towards your foot"), which would presumably represent
gir.zu.M, with no reduction of the fnal vowel, is also found. Such variation can occur
within one text, or even in different copies of one Sumerian literary work. For example,
line 72 of the Sumerian composition entitled "Schooldays" has the expression "towards my
hand", u.gul
O
.e. In most copies of the text, this is written: u-g!l
O
-M. However, at
least one copy has: u-g!l
O
-u.
The problem is further compounded by the fact that the e-sign also has a reading e,
so that if the terminative case-marker follows a word ending in lel, the writing is
ambiguous.
It is possible that the writings in -e should be understood as morphographemic,
standing for: I/. The scribe wrote the full, more original form of the morpheme, even
though in certain phonetic contexts it had been reduced in speech. That is, -e is written
conventionally for the terminative case-marker, without regard for its precise phonetic
shape. Other such morphographemic writings occur in Sumerian.
-Usage of -e
As stated above, the terminative case marked in -e in this fxed expression is
apparently not cross-referenced in any of the Ur III royal inscriptions. Curiously enough,
how
ever, there are a few cases in royal inscriptions from earlier periods where the
ten
inative in -e in such expressions is so cross-referenced by its dimensional-prefx.
This is stange, because the early texts are usually less explicit than the later texts. Also,
this dimensional-prefx often appears in the writing of Ur III administrative texts. Such a
dis
tibution indicat
e
s that factors such as genre and style help determine the use or non-use
of
the
dimensional-prefx .i (and of the use and non-use of the dimensional-prefxes in
g
eneral).
70 Lesson 4
For anyone who has studied Akkadian, the Sumerian terminative in Iel immediately
calls to mind the Akkadian terminative in liU. To some degree, the two morphemes
overlap both in form and in function. At least as far back as 1925, it was proposed that the
Akkadian terminative morpheme was borrowed from Sumerian (Albert Schott). However,
most scholars do not accept this view, for two reasons: First, the Akkadian terminative
appears to have cognates in other Semitic languages; it has recently turned up in Eblaite, in
certain limited contexts. Second, there do not appear to be any cases where Akkadian has
borrowed Sumerian grammatical morphemes. Similarly, it is difcult to believe that the
Sumerians borrowed the morpheme from the Akkadians. Although rare, it does show up
in very early Sumerian texts.
It is more likely that the two are independent developments. It is not impossible,
however, that the chance formal similarity between the two morphemes has caused the two
to infuence each other in meaning, pulling them closer together in meaning than they may
have been at some earlier period. Given the fact that Sumerian and Akkadian were in close
contact for over a millennium, such reciprocal infuence upon the grammar is not too
surprising.
- Compound verbs
Most compound verbs are of the type noun-verb, where the noun is (historically) the
patient of the verb. It is not sure whether these nouns are to be regarded as patients in
synchronic terms; this problem is discussed in Lesson 1 2. In some cases, the noun is in
one of the adverbial cases. However, more complicated compound verbs also occur, of
differing types: adjective-verb, noun-adjective-verb, and even noun-noun-verb. In the case
of the latter, one noun is (historically) the patient, and the other is (historically) in an
adverbial case. An example is "to pray", kiri
3
.u . . . W, literally, "to place (g@ the hand
(]) on the nose (kiri
3
)"; kiri
3
is either in the locative case (kiri.a) or the locative-terminative
case (kiri.e).
Because there is no obvious formal way to defne compound verbs, it is a legitimate
question to ask whether such a class of words actually exists. If more were known about
the etymology of each individual case, one might be less inclined to even posit the existence
of the class of compound verbs.
-Votive inscriptions
This text is a typical votive inscription, a second subclass of royal inscriptions as
distinguished by Hallo. Votive inscriptions are found on objects which were actually
donated and placed in a temple. The object can be of various kinds; this pacular
inscription is found on a stone bowl. Possibly, such vessels would have been used for
ceremonial food offerings. Since stone is not common in Sumer, such a vessel would have,
been considered as something special. In this book, the following votive objects occur: a
stone headdress (Text 1 1 ), a vase (Text 1 2), beads (Texts 1 6, 1 8a, and 1 9a), and a cylinder
seal of limestone (Text 22).
The purpose of such votive objects was to convey a hope from the donor for the long
Lesson 4 71
life of the king (the donor might or might not be the king himself). The objects were not
"functional", as we would understand the term; that is, this bowl was not actually used as a
daily eating utensi. Similarly, the votive cylinder-seal in Text 22 was probably not used as
a daily, routne, cylinder-seal.
Except for the component of the compound verb, there is no direct object (patient) in
the text. This i s because the votive object itself can be thought of as consttuting the direct
object (patient).
The use of the term "votive" to describe such inscriptions has been criticized by A.
Grayson:
The etymology of the word "votive" implies a vow and, since no vow is
involved in the ancient Mesopotamian texts under discussion, the term i s
incorrect. They are certainly not "votve" or "ex-voto" inscriptions in the
ancient Roman sense where a vow preceded the dedication ( 1980: 1 57).
For this reason, Grayson prefers to refer to such texts as "dedicatory inscriptions".
72
This is another brick.
Sign-list and vocabular
Utu Utu (ON, masc)
Notes
Lesson 4
Text 4a
supplemen tary
Utu This was the son of Nanna, and older brother of Inanna. He was primarily the god of
the sun (his name means "the [visible] sun"), but was also connected with tuth, justice, and
law-giving. He was equated with the Semitic Shamash, who is pictured on the top of the
stela of Hammurapi as handing over the law code written on the stela to Hammurapi. Utu
was especially worshipped in Larsa and Sippar.
Lesson 5
Lesson 5
This text is another building inscripton, inscribed on a clay cone.
Sig-list ad voabua
En-lil Enlil (DN, masc)
En-erin2-nun Enerinnun (canal name)
1
kur mountain; highland; foreign land
i d (7) river, canal
_ _ nidba food offering

ba-al to dredge, excavate


ra
t:
ka
Notes
73
En-IU Father of Nanna; the most important god of the Sumerian pantheon. His name
means "Lord Ai" or "Lord Wind", but he was in general responsible for the orderly
running of the universe (although he had a destuctve side as well).
Enlil functions as the active leader of the Sumerian gods; he has apparently displaced
the sky-god An fom this role. He was worshipped at many places, but his special
sanctury was the E-kur in Nippur.
The lil-sign is the same ke
4
-sign seen previously.
(The term en, "lord", is discussed in Lesson 9.)
En-ern
2
-nun The location is uncerain, as is the etymology (although it is presumably
Sumerian).
k The original meaning of this word was probably "mountain"; the kur-sign, in fact, i s
thought to be the picture of three mountain tops. The word then comes to mean "foreign
land".
id This sign is composed of two elements: I , which by itself represents , "water", and
_ which by itself represents ur, "watery deep, sweet waters" (see Lessons 1 and
14) , or Nammu the goddess. The original meaning of the sign may then have been
74 Lesson S
something like "water coming from the sweet-waters", as opposed to "water coming from
the salt-waters", and as opposed to "rain" (Mg, which is composed of the -sign followed
by the an-sign: T ).
ndba The reading is uncertain; it is also tansliterated as nindaba. When scholars are
unsure of the reading of a "compound logogram" (a single "word" graphically composed of
several individual logograms), they occasionally add in parentheses the reading of the
component parts. Thus, this sign is also tansliterated as: nidba (PAD-
d
INANA), or any
of several variants, such as: nidba (SUKUR2 _
d
INANA), since it is in fact not clear
exactly what all the components of this particular logogram are!
This word was borrowed into Akkadian as nindabu, although the forms nindabbu,
ndabbu. nidabu. nidbu and nidpu also occur. CAD tanslates the Akkadian word as:
"cereal offering, food offering, provisions".
ba-al This is the noral verb used to describe the restoration of a canal, that is, clearing it
by dredging it of accumulated silt and tash.
The verb is almost always written in this way, with two signs. There are a few
instances where it is spelled bal or ba-la. It is not sure what the writing ba-al implies about
Sumerian phonetics. The vowel may have been long, or there may have been a glottal stop
or a glide between the two / a/ -quality vowels. Because of this unusual writng, it has even
been speculated that the word is a borrowing from an as-yet unidentifed language. For
convenience sake, it will be tanscribed here as: ba-al.
Lesson 5
Text 5
=
.
fJ !

*
@
4

" T `"

1
75
76
1 :
2:
3:
4:
5:
6:
7:
8:
9:
10:
1 1 :
Transliteration
d
En-li l
lugal-kur-kur-ra
luw--ni
Ur_
d
Nammu
1 1 U
' ki
..gi - nm
S
-ma
luw-Ki-en-gi-Ki-uri-ke
4
--ni
mu-na-du
i d En-erin2-nun
i d-nidba-ka-ni
mu-na-ba-al
Commentary
Lesson 5
Transcripton Translation
Enlil For Enlil,
lugal.kur.kur.a(k) king of the lands,
lugal.anL(r) his king -
Urnammu Ur-Nammu,
lugal.Urim. a(k) the king of Ur,
lugal.KiengLKiuri.k.e the king of Sumer and Akkad -
e.anL0 his temple -
mu.na.(n.)du.0 he built;
id Enerinnun the Enerinnun canal,
id.nidba.k.ani.0 his canal (productive) of food -
mu.na.(n.)ba-a1.0 he dredged.
2. There ae two common plural formations of nouns in Sumerian. For animate nouns, the
plural is fored by a sufxed .ene. For example, "gods" is digir.ene, written Qgir-re-ne.
For inanimate nouns, the plural is formed by reduplication of the noun: kur-kur, "lands"
(This type of plural formation is also occasionally used for animate nouns.)
It is quite probable that such reduplicated plurals were phonetically reduced, but
because of the morpheme-bound nature of the script, such reduction does not normally
show up in the writing.
In Text 6, the plural noun is the second element of a genitive phrase:
lugal. [kur.kur] .a(k), "king of the lands:
'
.
9. id means "canal" or "river". The name of the canal fors an appositive.
10. A second appositive. The sufx .ani refers to the entire genitive phrase: "his [canal of
food offering]". The genitive phrase "the canal of his food offering" would be:
id. [nidba.ani].a(k). The sequence of third-person possessive-sufx followed by the geni
tive maker is /anak/, not the expected /anik/, so this would be written i d-nidba-na.
Discussion: structure
The structure of this text is:
[Enlil, lugal.kur.kur.a(k), lugal. ani] . (r)
[Urammu, lugal.Urim.a(k),
1 ugal.KiengLKi uri.k].e
[e.ani]. 0
mu.na.(n.)du.0
[id Enerinnun, id.nidba.k.ani] .0
mu.na.(n.)ba-a1.0
benefactive
agent
patient
verb
patient
verb
Lesson 5 77
-Writing system
The word for "food offering" illustrates a prblem not yet seen in the script. Its
pronunciation as Inidbal is given by various lexical lists, where it is spelled out syllabically
as:
ni-id-ba. It is thought to derive from ninda "bread, food" and ba "to divide, to ap
portion"; ninda would presumably be an incorporated direct object (patient) of ba. (It is
occasionally spelled ninda-ba.) The pronunciation as Inidba/ shows a phonetic reduction
of */nindabal Inidba/. The various spellings in Akkadian, mentioned above, refect both
the older and later S umerian pronunciations.
It is difcult, however, to say how this word came to be written as "PAD
d
INANNA". That is, there does not appear to be any way to phonetically relate the word
Inindabal or Inidbal to the individual signs forming this logogram. Rather, the reading
Inidbal refers to the "sum" of the three signs. Without the evidence of lexical lists, in fact,
there would prbably be no way to fgure out that this group of three signs was to be read
as Inindabal or Inidba/. The pad-sign has several different readings, and in Akkadian
stands for several different words: kusapu, "a kind of bread" (prbably of Semitic
etymology); kurummatu, "food portion" (a Sumerian word), etc. The signifcance of the
Inanna-sign (if that is how it is to be understood here) is uncertain.
-Proper names
In the bilingual texts frm Ebla, the equivalent of En-lil is given as: I-l-Iu. This
seems to agree with later Akkadian pronunciations of the name, which also show an
assimilation of IEnlill IIllil/; some Sumerologists, in fact, transliterate the two signs En
lil together as: Elli!. (It has also been speculated that the interpretation of the name as
"Lord Wind" is a Sumerian folk-etymology, and that the word is of pre-Sumerian
etymology.)
- Cones
The building inscriptions seen up to this point have all been inscribed on bricks. Text
5, on the other hand, was inscribed on what is commonly known as a "clay cone" or "clay
nail". Clay cones were used thrughout Mesopotamian history; their form and functon
varied to some degree from period to period. A detailed description is that of Grayson,
describing the clay cones of the Neo-Assyrian period. The clay cone
is an oblong conical object of clay. It is tapered almost to a point at one end
and at the other there is a large semi-spherical head. The same inscription
usually appears on both the shaft and head. The shaft was commonly
inserted in the upper portions of walls with the head, which was painted a
bright colour, prtruding ( 1 980: 145).
In this view, the clay cone would have been at least partially visible to on-Iookers.
Other scholars belive, however, that the prtruding end would have been plastered over,
covering up the inscription (at least, in the Ur III period). In fact, the exact purpose of these
co
nes is still something of a mystery to us.
Woolley found such cones in situ, forming part of the terrace of the ziggurat of Ur-
1
I0%n3
Nammu:
Such cones were familiar enough as objects on museum shelves, but now for
the first time we saw them in position just as the builders had set them four
thousand years before . ... One felt a quile unscientifc thrill at seeing those
orered rows of cream-coloured knobs which even the people of Ur had not
seen when once the tenaced wall was fnished and plasterd (1982: 140).
The latest such cones found HUr date from the Neo-Assyrian period. Curiously, they
were not Hthe wal1, but were buried below the foor.
Clay cones were usually inscribed in several duplicates; at least eight copics ofTcxt 5
are known. The point of such cones, again quoting Woolley, was not to "parade [the
ruler's] achievements before his fellow-men, but to keep the record of his piety fresh in the
mind of the god, who presumably can see through a brick wall" (1982:228).
Some scholars differentiate between clay "nails" and clay "cones". Gelb says that clay
nails "are easily recognized by their mushroom shape. with broad, thick heads and short
shafts. ... Clay cones are characterized by a total or almost total lack of the head"
(1948:267). Different places and periods seem to prefer one or the mher; it is also possible
they had different functions. Gelb says that "the function of such nails and cones is much
the same as that of tablets commemorating the erection of public Slctures in modem
times" (1948:268).
No photograph of Text 5 is available. The fOllowing is a photograph of a dedicatory
cone fm the rime of Gudea's father:
It was mentioned in Lesson 3 that it is not too uncommon for building inscrptions to
treat more than one activity. In the case of building inscriptions inscribed on clay cones, it
is less common for more than one acriyity to be mentioned. In this partcular text, the
Enerinnun canal may have brought the waters which irrigated the felds of the temple being
rebuilt.
All of the Ur HI kings were involved with repairing the canals and drainage systems
of Mesopotamia, by dredging and rebuilding. The names of many canals in use during the
Ur ID perio are known; most of the names a Sumerian. Urnammu is known to have
built and to have repaired a number of canals. It is usually assumed that much damage had
been done to these systems by the Guti, who are thought to b responsible H bringing
about the fall of the Dynasty of Akkad. Revisionist thinking, however, believes that the
Guti did not do as much damage as is commonly thought, nor did they hold that much
control over Mesopotamia. Most scholars seem to feel that inscriptons such as Text 5
refer more to routine maintenance and expansion of the canals; various kinds of
administative texts refer to such activity.
Lesson
This is another brick of Ur-Nammu.
Sign-list ad vocabulary
_ Determinative preceding objects of wood. Transliterated by a superscript "gi".
An An (DN, masc)
_ digi god

|||,!
kiri
6
garden

barag (bar

) dais
ki place, earth
-'
''

man to be splendid, magnifcent


Notes
sikil to be pure, clean
gub to stand; to make stand, to plant
gi In addition to its use as a determinative, gi means "tree; wood; object made of wood".
The Akkadian equivalent,
i
1, is translated by the CAD as: " 1 . tree, 2. timber, lumber,
wood, wooden implements, aromatic wood, frewood, 3. wooded area". It is sometimes
tansliterated as ge.
An The god of the sky. At one time he may have been the active leader of the Sumerian
gods, but at some point prior to our written records he was displaced in this role by Enlil.
Scholars sometimes refer to him as "shadowy", or as a kind of deus otiosus.
His name is almost always written without the divine determinative.
digi It is usually assumed nowadays that this word was pronounced with a velar nasal.
Some think it may have been pronounced I dingir/, and in fact it is most commonly
transliterated as dingi.
79
80 Lesson 6
g
Bkiri6 This is also transliterated with an initial / g/, especially in older transliterations:
giri
1 1
and giri
1
2 . Some Sumerologists do not think that the gi-component here is a
determinative, and so it is also transliterated: gi-kiri6 . (This is also the view found in
some lexical texts.) However, other Sumerologists believe that the kiri
6
-sign includes what
here is called a determinative; that is, the one sign kiri6 is composed of two separate ele
ments: _ and , and so there is no determinative. The most current practice is to
read the two signs as
gi
kiri
6
.
barag The PSD translates this simply as "dais"; many Sumerologists translate it as
"throne-dais". It was borrowed into Akkadian as parakku, which is translated by AHw as:
"Kultsockel, Heiligtum".
Text 6 was inscribed on a brick which apparently was part of the barag of the temple.
I It is not known what the cuneiform sign is a picture of.
mal This is not common in fnite verbal forms. The most common Akkadian equivalent
of its use as an adjective is Trl1, translated by the CAD as: "frst-rank (in importance,
quality), outstanding (in size), august, excellent (used only as a poetic term)".
si This is also not common in fnite verbal forms, but its adjective is frequent. The
usual Akkadian equivalent of the adjective is ellu, glossed by the CAD as: " 1 . clean, pure,
2. holy, sacred, 3. free, noble".
Lesson 6
81
Text 6
8
2
Lesson 6
Transliteration Transcripton Translation
1 : An lugal-digir-re-ne Anlugal.digir.ene. (k) For An, the king of the gods,
2: lugal--ni lugal.ani. (r) his king -
3: Ur_
d
Nammu Urammu Ur-Nammu,
4: lugil-Urim
s
ki
-ma-ke
4
lugal.Urim.ak.e the king of Ur -
5:
gi
kiri mah
-o-
kiri.mab

0 a magnifcent garden -
6: mu-na-gub mu.na.(n.)gub.0 heplanted;
7: barag ki-sikil-la barag.0 ki. sikil.a A dais, -in a pure place -
8: mu-na-du mu.na.(n.)du.0 he built.
Comenta
1. The sign can be very confusing in context. It can represent the deteninative before
divine names; the god An (as in Text 6); or the word digI, "god" (or the word an, "sky") .
. ene is the nonal plural marker for animate nouns. In the previous lesson, re
duplication occurred as the nonal plural marker for inanimate nouns.
5. mab is an adjective fom a verbal root. As discussed in Lesson 2, some of what are
translated as adjectives end in the nominalizer lal, e.g., kalag-g kalag.a Ikalaga/.
Other adjectives do not. Certain adjectives sometimes appear with the nominalizer lal, and
sometimes without it, but with no apparent difference in meaning.
The fons in .a are sometimes described as "passive participles", and those in .0 as
"active participles". For example, the adjective mab could be described as an "active
participle in .0", mab.0. This, however, may not be a valid use of the term "participle".
Although etymologically
gi
kiri
6
-mab is composed of two words, it may have been
"felt" as one word. It was borrowed into Akkadian as kirimabg, translated by the CAD as
"pleasure garden".
7. The I gl of barag is an amissable consonant, so the sign has both a long value (barag)
and a short value (bara2 or bara). Since in Text 6 Igl is word-fnal, some Sumerologists
would tansliterate it here by bara2.
sikil is another adjective from a verbal root. It is like mab, in that it appears without
the nominalizer I al.
The lal which fons part of the la-sign is a new case-marker, that of the "locative"
case. This case is used to express location, either spatial ("in that place") or temporal ("on
that day"). Spatially, it can usually be translated by the English "in".
Its basic fon is la/. Following a vowel, it usually does not contract. "In the earth",
for example, is almost always written: ki-. After a consonant, it is nonally written by a
sign which reduplicates the fnal consonant of the previous word. Thus, sikil.a is written
sikil-la.
8. The dimensional-prefx which cross-references the locative is .ni. In a sequence of
dimensional-prefxes, it always occurs last. In this line, one might have expected to fnd:
mu-na-ni-(n-du, but the dimensional-prefx does not appear. In fact, in the Ur III royal
Lesson 6 83
inscriptions, the locative .a is usually not cross-referenced. This is probably more than just
orthographic problem. In Lesson 4, an instance occurred where the dimensional-pref
for the terminative was not expressed. There, it was said that it may have been because of
the idiomatic nature of the phrase. However, such an explanation hardly fts the numerous
cases where the locative case is not cross-referenced. Perhaps locative phrases in general
were felt as less closely bound to the verbal phrase than were the other adverbial cases.
The logical antecedent of the datival dimensional-prefx .na is the nominal phrase in
lines 1 -2 of this inscription, which is part of a different, independent sentence. The same
situation occured in Text 3, where a datival dimensional-prefx referred logically to a
nominal phrase in a preceding sentence.
Discussion: structure
Te stucture of this text is:
[An, lugal.digir.ene. (k), lugal.ani] . (r)
[Umammu, lugal.Urim.ak] .e
[kiri.mab]

0
mu.na.(n.)gub.0
[kLsikil] .a
mu.na.(n.)du.0
-Amissability
benefactive
agent
patient
verb
place
verb
It may be useful here to summarize the ways the amissable consonants are refected in
the writing system. In the case of grammatical morphemes (which are normally written
syllabically), such as the genitive .ak, the amissable /k/ does not show up in word-fnal
position: "King of Ur" is written: lugal-Urim,
t
ma. When not in word-fnal position, it
does show up: The same expression, with an ergative case-marker, is written: lgal-
U
ki
k nm
S
-ma-c.
In the case of lexical morphemes (which are normally written logographically), it
cannot be determined from the writing system whether the Auslaut was pronounced or not.
That is, just by looking at the sign, there is no way to tell whether the word for "dais" was
read /barag/ or /bara/.
When the amissable consonant of a lexical morpheme is not word-fnal, there is less of
a problem in understanding the phonology, but still a problem in understanding the nature
of the orthography. As discussed in Lessons 2 and 4, should a form such as "for the sake
of his life" be understood as nam-ti I-la-ni-e or as nam-!-la-ni-se? Or should the adjective
for "mighty" be understood as kalag-g. or as kal a-gf!? Phonetically, probably all Sume
rologists would understand these forms to be /namtilange/ and /kalaga/. The question,
rather, is how does Sumerian orthography represent these pronunciations.
- Loan words
In a number of early loan words from Sumeri an into Akkadian, fnal voiced stops
appear as unvoiced, and usually as geminated: isib ("kind of priest", Lesson 19) ) mpu;
M ("an offcial") ) abarakku. Word-initial voiced stops usually appear as unvoiced in
84 Lesson 6
Akkadian: barag ) parakku.
It is not sure what this tells us about Sumerian phonology. As was discussed above,
the difference between the two sets of consonants Ib d gl - Ip t kl may not have been one
of voice.
-Locative
Locative phrases such as "in the earth", ki.a, are almost always written ki-. That is,
the lal of the locative case-marker almost always appears in the writing. It has been
speculated that the lal of the locative case-marker actually assimilates into or contacts into
a preceding vowel, but the script is morphographemic, and writes the lal anyway.
The locative case-marker .a does not usually appear when the head noun is the frst
element of a compound verb (kiri
3
.u ... gal was mentioned in Lesson 4). It is possible that
this is more than a case of assimilation or contraction. As discussed in Lesson 4, it is not
sure how "present" case-markers were in the case of compound verbs, in a synchronic
sense; they may have been deleted or somehow reinterpreted.
Although in the Ur III royal inscriptions the locative case is usually not cross
referenced by a dimensional-prefx, in the contemporaneous Ur III administrative docu
ments it quite frequently is cross-referenced. As was mentioned in Lesson 4, this may
mean that factors such as genre and style must be taken into account, to describe and
explain the different distributions.
-Adjectives
In Text 6 the adjectives man and sikil occur. In Text 2 the adjective kalag-g occurred.
As stated above, some adjectives end in .0, others in .a. Some adjectives are occasionally
found sometimes in .0, other times in .a. This situation is not well understood. The most
recent discussion is by Joachim Krecher ( 1 978). He believes that, at least in certain cases,
the forms in .a mark a nominal phrase as "defnite" or "determined" in some way; those in
.0 are the unmarked forms.
The diffculty in investigating this problem (and many other problems in Sumerian
grammar) is that it is not easy to fnd sentences which are very close in structure, but differ
only in the presence or absence of .a on an adjective. There are usually too many variables
involved, to be able to sort them all out.
Lesson 7
Two copies of this inscription are reproduced. The varation in the shape of the
cuneifon signs is fairly minimal. Text 7a is a stone foundation tablet. Text 7b is a brick.
The inscription appearng on them is a standard inscription, a further sub-class of royal
inscription as distinguished by Hallo.
Sign-list ad vocabular
1t man
Notes
Id According to Gelb, "The Sumerian word 1t is a noun meaning 'person, man
'
(in the
sense of homo, Mensch, not vir, Mann), and may be used for both males and females"
( 1 979b: 51). Jacobsen says that 1t "denotes a man (Akkadian awIlum) or woman (A
kadian aWIltum) who heads a household, f, or city" ( 1 987: 1 30 n. 17).
b

07
Text 7a
Lesson 7 87
Text 7b
88
1 :
2:
3:
4:
5:
Transliteration
Ur_
d
Nammu
1 l U
ki
..g.- nm
S
-ma
lugal-Ki-en-gi-Ki-uri
lu _
d
En-1il-la
in-du-
Comenta
Lesson 7
Transcription Translation
Urammu Ur-Nammu,
lugaI.Urim.a(k) the king of Ur,
lugal.Kiengi.Kiuri.(k) the king of Sumer and Akkad,
lu e.Enlil.a(k). 0 the man who built the temple
of EnliI.
Ln.du.0. a
4. _
d
En-lil-la e.Enlil.a(k). Following certain words which end in Ill, the la-sign (in
stead of the la-sign) is used to express the combination of the reduplication of the previous
word-fnal III with the lal of the genitive marker, or the lal of the locative case-marker.
Such a word is liI. There are other words ending in III where the la-sign (and not the la
sign) is used. In Lesson 6, there occured: ki-sikil-la. These differences in writing are a
clue that Sumerian had more than one type of Ill-sound. Not enough is yet known about
Sumerian phonology to say exactly what this means.
Lines 4-5 correspond to a relative clause in English, "the man who built the temple of
Enlil". In Sumerian, relative clauses (for want of a better term) are composed of two
elements: a noun serving as relative marker (corresponding in function to an English
relative pronoun), and a nominalized sentence, which stands in apposition to the preceding
relative marker. In Text 7, the relative marker is lu. The nominalized sentence is formed by
the addition of the nominalizer . a to a complete sentence. For example, "he built the temple
of Enlil" is: e.Enlil.a(k).0 mu.(n.)du.0. This can be nominalized by the addition of .a:
[e.EnliI.a(k).0 mu.(n.)du. 0] . a, meaning something like "(the one) who built the temple of
Enlil". Put into apposition with the relative marker lu, this means "the man who built the
temple ofEnlil".
The simplest way to understand relative clauses in Sumerian is to think of them as
being equivalent to big adjectives. For example, "the mighty man" is: [nitabHkalag].a.
"The man who built the temple of Enlil" is: [luHe.EnliI.a(k). 0 mu.(n.)du.0]. a. (As a
matter of fact, many linguists consider adjectives in general to be "reduced" relative clauses:
"the mighty man", in English, derives, in some sense, from "the man who is mighty".) The
principle is the same; adjectives and relative clauses fulfll the same function.
Sumerian does not have a morphologically distinct class of "relative pronouns".
Because of the frequency of the noun lu in these contexts, however, it functions almost like
a relative prnoun. The Akkadians themselves sometimes translated lu as a, the Akkadian
relative pronoun. Occasionally, other words may function as the equivalent of a relative
pronoun.
Although such a construction as: lu e.EnliI. a(k). 0 mu.(n.)du. 0.a is theoretically
possible, in fact the verb form used in Text 7 is different in two ways from the rather stock
or formulaic verbs seen in the previous texts. First, it uses a different conjugation-prefx.
The conjugation-prefx which occurred in all the previous inscriptions was Imul, written
mu. However, this text uses the conjugation-prefx lit. If not immediately preceded by a
Lesson 7 89
con
son
ant, this conjugation-prefx is nonally written by the i-sign (that is, the sign which
ocurred earlier with the value ni).
Second, here the personal-afx 1nl cross-referencing the agent is expressed in the
script. The in-sign expresses the combinaton of the .i conjugation-prefx with the .n of the
personal-afx. This particular use of the in-sign is quite regular.
It is diffcult to say why the fnite verb fons seen previously use the conjugation
prefx
!l, but the nominalized fon uses the conjugation-prefx i. As discussed in Lesson
1, the essential difference between mu and i is elusive. Jacobsen says that the conjugation
prefx i "presents the occurrence denoted by the verb as touching on the subject without
inwadly conditioning him in any lasting manner" ( 1 965:25 1). mu is the
mark of location of the occurrence denoted by the verb on the inside border
(.u) of the area of the speech situaton (m.) . . . . It adds to this implications of
emotional involvement of the speaker, of his being personally engaged
( 1965: 254).
However, not all scholars are as certain as Jacobsen in their conviction; Poebel, for
instance, seemed to have the opposite view of the relationship between mu and i. J.N.
Postgate has expressed perhaps the most negative view:
For many years a vexed question in Sumerian has been the distinction
between the prefxes mu- and i-, and our failure to defne the difference in a
satisfactory way has epitomized our helplessness before Sumerian grammar
as a whole ( 1974:24).
And, as discussed in Lesson 1, in actual practice the conjugation-prefxes are basically
ignored in tanslation.
It is likewise difcult to say why the personal-affx 1nl is written here, whereas it did
not appear in any of the previous fnite verbal forms. In the previous texts, the fnite verb
fon always had an expressed subject (agent). In this particular text, there is no expressed
agent, since the verb is inside a relative clause; the h of line 4 is the relative marker, not an
agent of the sentence as a whole. One might hypothesize a rule such that "expressed agents
do not use the personal-afx 1nl, but verbal forms without expressed agents do", but from
other texts it is known that the situation is not nearly as simple as this.
To sum up, II of line 4 is the relative marker of the relative clause. The relative clause
is foned by nominalizing a fnite sentence, by use of the nominalizer .a. The nominalized
clause stands in apposition to the relative marker.
Discussion: orthography
In both copies of this text, the relative clause is split into two cases. The frst case
contains the more nominal component, and the second contains the more verbal component.
It is not uncommon for long relative clauses to be split into two or even more cases.
The Sumerian word for "king", ilgal, is a noun-adjective compound from Il, "man",
and gal, "great, big". In older forms of the ilgal-sign, the gal-component was written a
little above and to the right of the Il-component. At times the two signs were totally
separated, and can even be written on two different lines of one case. As the cuneifon
signs gradually became more linear, the gal-sign shifted position, and so in "standard"
90 Lesson 7
Sumerian the ilgal-sign is all one sign, with the gal-component in front of the hl
component.
As will be discussed in Lesson 1 3, the order of cuneiform signs within a case in the
earliest texts was to some degree free, with the order-as-written not necessarily refecting
the order-as-read.
-Conjugaton-prefxes
As hinted at above, the fact that the conjugation-prefx 1 appears in the relative clause
instead of mu raises several obvious questions: Does 1 appear outside of relatve clauses?
Does mu appear inside of relative clauses? What about the distribution of 1 and mu in
general? What about different types of relative clauses: clauses where the relatve clause
modifes an agent ("the man who built"), clauses where the relative clause modifes a
patient ("the house which the man built"), etc. What about the distribution of the
con j ugation-prefxes in topic ali zed and emphatic sen tences?
Unfortunately, not all of these questions can be answered. The data are both
ambiguous and limited; the number of attested relative clauses is not that large. Without
access to native speakers, such questions cannot be tested.
In the Ur III texts, the verb form in the relative clause in this particular expression
seems always to appear as: in-du-. In some bricks of Gudea, the form mu-na-du is used
in a sentence as a main verb, while the verb of an embedded relative clause uses the form
in-du-:
d
Nin-gi-zid-da digi-ra-ni
Gu-de- ensi2 La
ki
III E-ninnu
d
Nin-gir-su-ka in-du-
Gir-su
ki
-ka-ni
mu-na-du.
For Ningishzidda, his god -
Gudea, the ensi of Lagash, the man who built the Eninnu of Ningirsu -
built his Girsu temple.
The most recent investigation of the conjugation-prefix lil is by Herman Vanstiphout
( 1 985). He believes that the oppostion mu vs. i cannot be understood simply on the level
of the individual sentence; rather, the larger context (discourse) must be examined. His
tentative results are that
Ii/ seems to carry substantially 'secondary' information in discourse
(accompanying or descriptive information, including consecutive verbs) . . .
on the supra-sentential or discourse level the prefix Ii/ serves as a back
gounding device ( 1 985: 1 1 ; 1 3).
He does not examine relative clauses specifi cally, but such an investigation might prove
useful.
It has also been suggested that 1 was a nasalized vowel, and that the presence or
absence of a following ! is a question of phonology, not of morphology or syntax. But it
is not even sure that nasalized vowels existed in Sumerian.
Lesson 7 91
-Standard inscriptions
It will be noticed that there is no fnite verbal fonn (or any other predicate) in this text!
Instead, there is only a personal name, followed by a series of titles or epithets. Jacobsen
has referred to such sentences as "label sentences". Hallo, in his discussion of the different
categories of royal inscriptions, refers to such inscriptions as standard inscriptions or
propery inscriptions. By "standard", Hallo means that the text is something like a fag or
other identifying device: It identifes the building as being the property ofUr-Nammu.
Such inscriptions, which serve to indicate ownership of a building, are often regarded
as the "simplest" fon of royal inscriptions. They usually consist of a royal name,
followed by a limited number of epithets, one of which may be a relative clause, as in Text
7a/b. There is no verbal predicate, These texts can be very short; Text 1 3c is a standard
inscription consisting of only two lines.
-Foundation deposits
The building inscriptions presented here up to now have been inscribed on either
bricks or clay cones. Text 7b is such a brick. A third category of building inscriptions is
referred to as "foundation deposits". These were actually buried under the foundations of
walls in a building, in a small pit. Text 7a is one sub-type of such foundation deposit, a
stone foundation tablet. In Woolley's words:
Foundation-deposits are found in the corners of buildings. Built into the
wall-foundations there is a small box of burnt bricks, lined with matting and
waterproofed with bitumen; in it is set a copper fgure of the king modestly
represented as a labourer carrying on his head a basket of mortar; at his feet
is a stone tablet in the fan of a pIano-convex brick; on the brick and on the
king's skirt is an inscription recording his name and that of the temple
( 1982: 1 61).
Occasionally, the fgurine or the tablet is uninscribed. It is thought that these
foundation tablets were intended to represent "model bricks".
The following photograph is of a fgurine which was found in Nippur along with Text
7a, and so bearing the same inscription. (At least two of these brnze canephore fgurines
are known.) The fgurine is thought to represent Ur-Nammu himself, carrying a basket on
his head with the building materials used to make the "frst brick" of a building.
L
n
7
97
Lesson 7 93
Text 7c
supplementary
Another brck.
Lesson
This text was inscribed on what is commonly called either a door socket or pivot
stone.
Sign-list ad vocabulary
_ i S-kur Ekur (temple name [TN)
ki. . .igi2 to love
Notes
E
-kur This was the main temple of Enlil in Nippur, by far his most important sanctuary.
It was in the Ekur that the assembly of the high gods would meet, as occasion demanded.
The god An presided over these meetings, but it was the responsibility of Enlil to cary out
the decisions. One of the reasons meetings were held was to select the rulers of Meso
potamia.
ki. . .ag

The verb meaning "to love" is written both and m. Most
Sumerologists believe that both writings represent /ki . . . aga/. The frst writing is to be
understood as: ki . . . i2
'
and the second as: ki . . . M-g. This means that the sign has
two values: ig and igi2. The problem of the "overhanging vowel" in such cases will be
discussed in Lesson 1 1 .
It is not clear what the two elements of this compound verb mean. The most common
meaning of igi2 is "to measure" (the Akkadian equivalent is madadu), but it can also mean
"to mete out". The most common meaning of ki is "earth". It is hard to say how an
expression like "to measure the earth" could come to mean "to love". Either the word ki,
"earth", had some other meaning, not known to us, or else the element ki in k
L.
M
i2 is an
entirely different root ki (that is, a homonym).
The most common Akkadian equivalent ofki. . .igi2 is ramu, "to love".
95
96 Lesson S
Text 8

+
P
~
$


-
"

7

-
1 :
2:
3:
4:
5:
6:
Transliteration
d
En-lil
lugal-kur-kur-ra
lugal-i-ni
Ur_
d
Nammu
nitab-kalag-gi
I I U
ki
jg..- nm
S
-ma
Lesson 8 97
Transcription Translation
Enlil For Enlil,
luga1.kur.kur.a(k) the king of the lands,
luga1.ani.(r) his king -
Urnammu Ur-N ammu,
nitab kalaga the mighty man,
luga1.Urim.a(k) the king of Ur,
7: lugr-Ki-en-gi-Ki-uri-ke4 luga1. Kiengi.Kiuri.k.e the king of Sumer and Akkad -
8: :-kur Ekur the Ekur,
9: -ki-g-g-ni e.ki.aga.a.ani.0 his beloved temple -
10: mu-na-du mu.na.(n.)du.0 he built.
Commentay
9. In Text 2, the adjective kalag-gi occurred, derived from the root Ikalagl with the
nominalizer .a. In Text 8, the same formation occurs, but from a compound verb. ki. . .ig

is "to love"; the adjective "beloved" is [ki. aga] . a. The scope of the nominalizer here
involves both the nominal and verbal elements of the compound verb; therefore, the
nominalizer is given in tanscription, unlike the case with "simple" adjectives such as
Ikalaga/.
As discussed in Lesson 1 , the third person possessive-suffx after vowels appears
both as i-ni and as ni. Perhaps, then, this form should be understood as: ki.aga.a.ni,
instead of: ki.aga.a.ani.
Discussion: structure
The structure of this text is:
[Enlil, luga1.kur.kur. a(k), luga1.ani]. (r)
[Uramu, nitab.kalaga, luga1.Urim.a(k),
luga1.Kiengi.Kiuri.k] .e
[Ekur, e.ki.aga.a.ani] . 0
mu.na.(n.)du.0
-Door sockets
benefactive
agent
patient
verb
Many texts inscribed on door sockets have been preserved; for a photograph of such a
stone, see Text 17a below. Such stones were partially underground, and were used to hold
a door. In Woolley
'
s words,
The Sumerian door consisted of a wooden leaf fxed to a pole rather higher
than itself; the projecting top end was held by and revolved in a metal ring
attached to the lintel, the lower end was shod with metal and went down
through a hole in the pavement to rest and turn on the hinge-stone. This was
98 Lesson S
a boulder of (imported) hard stone, limestone or diorite, in which a cup
shaped hollow had been cut to take the pole-shoe, and generally one part of it
had been smoothed and inscribed with the name of the king who dedicated
the building and of the god in whose honour he built it. . . . Imported stones
were valuable and an old stone would often be taken away and re-used for
some building other than that for which it had been intended, so that the old
inscription no longer applies ( 1 982: 1 60- 1 61 ) .
I some cases, door sockets were re-used for reasons of piety; in other cases, i t was
purely for reasons of economy.
As Woolley's description implies, the building inscription inscribed on the door socket
would not normally be exposed to view. Royal inscriptions in general were not meant to be
seen by contemporaries of the builder, but rather by future rebuilders.
In Hallo's scheme, door sockets are a subdivision of building inscriptions, which are
themselves one of the major divisions of royal inscriptions.
Lesson S
99
Text 8a
supplementay
Another brck.
Lesson 9
This text is another brick.
Sign-list ad vocabular
Unug (Unu) Uruk (ON)

'

E-temen-ni -guru
3
Etemenniguru (TN)
_ dumu son
_ sag head
en lord

g to retur, to restore
bi, be
Notes
Unug One of the most important cities in southern Mesopotamia; it often played a role in
political history. Before becoming king, Ur-Nammu had been military goveror of Ur
(akkana-Urim
5
kC
ma) under the control of Utu-Hengal in Uruk.
The etymology of the name is unknown; this is discussed below. The pronunciation
of the name as IUnugl is known from syllabic writings. However, Semitic spellings show
Ir/: Biblical Erech, moder-day Arabic Warka, etc. ; the lexical lists also give the Akkadian
equivalent with Ir/: U-ru-uk. It is not known why the Sumerian form shows 1nl while
the Semitc fors show Ir/.
E
-temen-ni-guru
3
This is the name of Ur-Nammu
'
s ziggurat at Ur, illustrated in Lesson
1 . The etymology is uncertain. It is sometimes tansliterated as E-temen-ni -ila2'
sag This is literally "head". However, it not infrequently forms the second element of
noun-noun compounds, where, as in many languages, it can take on metaphorical uses.
For example, ka-, "beer-head", is "top quality beer"; dumu-, "son-head", means
"eldest son".
has several Akkadian equivalents: reu, "head" and several derived meanings;
q.! qqadu, also "head" and several derived meanings; amelu, "man", lu, "young man", etc.
en This is norally tanslated "lord", a purely conventional translation. The ruler of Uruk
is always called an en. Its meaning is further discussed below.
1 01
102 Lesson 9
Text V
1 :
2:
3:
4:
5:
6:
7:
8:
9:
Transliteration
d
Nanna
dumu-
d
En-lil-hi
lugal--ni
Ur_
d
Nammu
nitab-kalag-
en-Unug
kC
g
I I U
ki
..g- nm
S
-ma
lugal-Ki-en-gi-Ki-uri-ke
4
1 0: E-temen-ni -guru
3
1 1 : -ki-g-g-ni
12: mu-na-du
1 3: ki-be mu-na-g4
Commentar
Lesson 9 10
3
Transcr ption Translation
Nanna For Nanna,
dumu.sag the eldest son
Enlil. a(k) of Enlil,
lugal.ani. (r) his king -
Urnammu Ur-Nammu,
nitabkalaga the mighty man,
en.Unug.a(k) the lord of Uruk,
lugal. Urim. a(k) the king of Ur,
lugal.Kiengi.Kiuri.k.e the king of Sumer and Akkad -
Etemenniguru the Etemenniguru,
e.ki.aga.a. ani . 0 his beloved temple -
mu.na.(n.)du.0 he built;
ki . bi.emu.na. (n.)gi.0 he restored it to its proper
place.
13. ki-be mu-na-g4 means "he restored it to its place", that is, rebuilt the temple at its
original location. This entire expression was borrowed into Akkadian as: ana aSrBu itOr,
"he retured (it) to its place" .
. bi is the inanimate possessive-suffx, "its", referring back to the temple. It is the
inanimate equivalent of the animate possessive-suffx .ani.
The most obvious interpretation of the frst two words of this line would be to read ki
bi, interpreting it as the direct object (patient) of the verb g4: "he retured its place".
However, there is evidence that ki-bi here is not a direct object (patient), because variations
of this formula already include an expressed patient. For example, -ni ki-be mu-na-g4,
"he restored his temple to its place"; bad-bi ki-be
4
mu-na-g4, "he restored its wall to its
place".
One interpretation is to read ki-bi, for ki. bi. (se), that is, a terminative nominal phrase.
The terminative .Se becomes IS! afer a vowel , and since Isl is an amissable consonant, it
would not appear in the script. While it is true that the morphology of the terminative case
marker is not entirely clear, one might expect to fnd at least one case in early texts of the
terminative appearing in this formulaic phrase, but none apparently occur; note that it
regularly occur

after a vowel in forms such as nam-til-Ia-ni-se. It is true that in later texts
ki-bi-se occasionally occurs in somewhat similar contexts; however, it is possible that this
is due to infuence from the Akkadian equivalent.
The second possibility is to read ki-be, for ki.bi.e. The .e is the marker for a case not
yet seen, the "locative-terminative" case. It is diffcult to pin down a specifc function for
this case. It shares some of the characteristics of the locative case, and some of the
10 Lesson 9
characteristcs of the terminative case. In this particular fxed expression, the force
conveyed by the locative-terinative is: "He returned the temple to its original place". That
is, its meaning here is close to that of the terminative marked by -e.
The form of the locative-terminative dimensional-prefx is much open to dispute. It is
often not cross-referenced at all in the verbal prefx chain. Other times, it seems to use the
dimensional-prefx corresponding to that of the locative case, that is, .ni. (According to
some Sumerologists, .ni in such contexts is for *n-e; the .e is the dimensional-prefx of the
locative-terminative, and the .n is an optional pronominal-prefx, discussed in Lesson 1 5. )
In this particular case, it is not cross-referenced at all in the verbal chain.
The . 0 cross-references the patient. There is no expressed patient in this sentence; the
logical patient is the nominal phrase in line 1 1 . Similarly, the logical antecedent of the
datival dimensional-prefx .na is the datival nominal phrase in lines 1-4.
Discussion: structure
The structure ofthis text is:
[N anna, dumu.sag.Enlil.a(k), lugal.ani] . (r)
[Urammu, nitab.kalaga, en.Unug.a(k),
lugal.Urim.a(k), lugal.KiengLKiurLk] .e
[Etemenniguru, e.kLaga.a.ani] . 0
mu.na.(n.)du.0
[kLbi] .e
mu.na.(n.)gL0
-Sign formation
benefactive
agent
patient
verb
place
verb
Over time, the Sumerian writing system produced a number of new cuneiform signs;
the history of these developments is rarely visible to us. One such process was the addition
of short strokes to an already existing cuneiform sign, thereby modifying its meaning. For
example, the word sag, "head", is represented by a sign which was originally the picture of
the head and upper torso of a man:
x. C:
-f '
The word ka, "mouth", is represented by the
same sign, but with the addition of short strokes over the region of the mouth: l. :r .
Akkadian scribes referred to these extra strokes as gunO. This is an adaptation of a
Sumerian word gunu, apparently meaning "colored". Modern scholars will sometimes
refer to, e.g., the ka-sign as "SA
G
+gunG". This convention is sometimes necessary, when
the value of a "gunOfed" sign is not known.
- Functions of .e
The locative-terinative case is marked in .e, and the agentive-ergative case is marked
in .e. An obvious question which springs to mind is, are the two related? It is probably no
accident that the two case-endings share the same phonological shape, / e/. The parallel has
been made to the English preposition "by". This can express a locative ("by the river"), an
instrumental ("by the hammer"), or an agent ("by the man"). The agentive marker .e may
have developed out of this locative-terminative .e. This .e may have started to lose some of
its functions, which began to be taken over by the locative case in .a and by the terminative
Lesson 9 105
case in .e. Haayer has discussed this point, in terms of universal tendencies in language:
One of the most characteristic features found in case marking in ergative
languages is that the ergative case is often identical with another case, most
often the genitive or instrumental, sometimes the locative or dative. In
Sumerian, for instance, the ergative case is marked by the postposition -e,
which is identical to the locative-terminative -e, and is in origin a deictic
pronoun ( 1986: 80).
Although we understand and translate simple Sumerian sentences such as lugal.e e.0
mu.n.du.0 as "the king built the house", it is usually assumed that at some "Proto
Sumerian" stage, the meaning may have been something like "a house got built, connected
with the king", or "there was a building of a house by the king". That is, to some degree,
pre-historic Sumerian (and historic Sumerian?) should be understood as basically "passive"
in nature. It has been said that in ergative languages, the patient of the sentence is the
"topic" of the sentence (while the agent is the "comment"), but in accusative languages, it is
the subject which is the topic (while the patient is the comment).
-Noun compounds
dumu-sag is literally "son-head". This represents a case of noun-noun compounding.
This is not a productive method of word formation in historic Sumerian, but a few such
cases exist. Several early proper names are noun-noun compounds:
d
En-H, "lord-wind";
E-kur, "temple-mountain", etc. ; these are not genitive formations.
The element nig, used to form concrete (and occasionally abstract) nouns from verbal
roots, is in origin a noun meaning "(some)thing". The original meaning of nam is less sure,
but possibly had a similar origin, or may have meant something like "state-of-being".
Thus, abstract and concrete nouns formed from either D or nam originally represented
noun-noun compounding.
-History
The different functions of the en and l! gal have been much discussed; they varied to
some degree from place to place and from period to period. In Jacobsen
'
s seminal article on
"Early Political Development in Mesopotamia", he stated that in the earlier periods the en
(Akkadian blu) was more of an "administrator" while the l! gal (Akkadian aru) was a
"warleader":
In the case of the en the political side of the offce is clearly secondary to the
cult function. The en
'
s basic responsibility is toward fertility and abundance .
. . . The "king", l! gal, in contrast to the en was from the beginning a purely
secular political fgure, a "warleader" ( 1 957 : 375 n.32).
As for the more "original" meanings of the telms,
The Sumerian term en which is generally translated "lord" denotes basically a
productive manager, someone with magic gifts to make things thrive
( 1 987: 20 n.2; cf. p. 277) . . . . Under the early political forms . . . the king
(l! il ) was usually a young man whose task it was to lead the army in war
( 1 987:236 n.4).
106 Lesson 9
-Substrate
The etymology of the name Uruk is unknown. Many of the oldest cities in Sumer
have names which are not apparently Sumerian. Such names go back to the language(s)
spoken by the people(s) living in southern Mesopotamia before the Sumerians arved; the
name of the city of Ur may be one of these names. Gelb says:
Almost all the Mesopotamian geographical names found in the earliest
Sumerian sources are non-Sumerian and non-Akkadian and must be
assigned to the proto-population of Mesopotami a. This conclusion is true of
the names of rivers and mountains, as well as of cities and countries. Only in
the Pre-Sargonic period do we fnd the frst attestation of Sumerian
geographical names ( 1 962:49).
(A number of these place-names are discussed in Limet 1 97 5b. )
These substrate peoples, about whom virtually nothing is known, were referred to by
Benno Landsberger as "Proto-Euphrateans"; they are also sometimes referred to as "Proto
Tigridians". The words for certain material objects and certain professions in Sumerian go
back to this language, for example, nagar, "carpenter", which has no obvious Sumerian
etymology. Some of these substrate words then passed on to Akkadian, and eventually on
to Aramaic, Hebrew, and Arabic.
Scholars disagree in their views as to how much of Sumerian vocabulary is of
substrate origin. Some are inclined to see a large number of substrate words in Sumerian,
including many place-names and divine-names; other scholars are less convinced.
Jacobsen, for instance, has Sumerian etymologies for several city-names, which other
scholars regard as substrate names. There is rarely enough evidence to decide any
paticular case.
- History
Contol of Uruk was important to all the Ur HI kings. Hallo says:
Certainly the two cities [Ur and Uruk] had a venerable history of dynastic
and administrative union behind them . . . . Ur under Ur-Nammu was heir to a
long history of dynastic and administrative union with both Uruk and
Lagash . . . . Nippur is the religious center, Ur the political capital and Uruk,
fom all indications, the ancestral home of the dynasty ( 1 966: 1 36- 1 38).
- History
It i s diffcult to say how literally the expression "he restored it to its place" should be
understood. The Sumerian phrase (and the corresponding Akkadian phrase) is somewhat
ambiguous; it can mean either "to restore to a former spot" or "to restore to a former state".
Kings of the Neo-Babylonian period specifcally claimed that they took care to rebuild
temples exactly on old foundations. Woolley says that
It was customary in Mesopotamia, when rebuilding a temple, to incorporate
the earlier one within the core of the platform upon which its successor was
to be set. This ofen meant largely di smantl i ng it ( 1 982: 1 09).
Lc onV I!
Nabonidus. for example. has left several inscriplions in Ur, in which he claims to have
restored the ziggurat of Ur-Nammu (the Eterenniguru). He states. in fact. that Ur-Nammu
started the work on the ziggural, but did not fnish it; Ur-Nammu's son and suc",essor
Shulgi also worked on the complex, but did not fnish it; only he. Nabonidus himself,
completely fnished and restored it. The following drawing is WooJley's reconstucton of
Nabonidus' ziggurat; it is instructive to compare i t with Woolley's reconstuction of Ur
Nammu's ziggurat, given i n Lesson i. Ur-Nammu's ziggurat itself was built over an earlier
temple, which itself was built over an even earlier temple.
L LLLKP L AbLPLL KLLK
.. , PO 8C
_. - .. ..
108 Lesson 9
As another example of temple-rebuilding, the Ishtar Temple of Assur was in existence
some two thousand years, and was fequently rebuilt. Richard Ellis says that "It was
always in about the same place, though sometimes the new version would b placed to one
side of the earlier ruins" ( 1968: 1 2). The temple of Inanna at Nippur had an even longer
history. It was built and rebuilt from at least as early as Early Dynastic I to late Parhian
times -that is, about 270 BC to 1 50 AD. There were at least eleven major building levels.
The temple varied in size fom perio to perio, usually getting bigger. The new
sanctuaries were normally built over the previous ones.
The principles behind the orientation of Mesopotamian temples are not at all clear,
especially in the older periods; some of the evidence seems contadictory. Nor are the
means by which the Mesopotamians determined the orientation known. Ganther Matiny,
writing in 1940, says that
Astonomical orientation is ... especially noticeable in the case of late
temples. The direction of orientation should probably b understoo as the
direction in which the go
'
s statue faced . . . . In Neo-Babylonian times
orientation based on individual stars assigned to specifc deities came into
vogue (1 940:92).
Sally Dunham, however, writing almost ffty years later, is less sanguine:
Very little is known about how the ancient Mesopotamians oriented and
measured off the ground plans and precincts of their temples, although we do
know such measuring was important enough to b mentioned in their royal
inscriptions and religious texts . . . . Still toay nothing is known about i and
how the ancient Mesopotamians used astonomy to orient their temples
( 1 986:39, and n.37).
Matiny thinks that the "Gimisin (Le., Shu-Sin) Temple" in Eshnunna was oriented
toward the city ofUr:
Exactly along the projected axis of the Gimilsin Temple in the direction in
which the go's statue faced, at a distance of about 30 k. toward the
southeast, lies Ur, the residence of Gimilsin. Is it possible that the deifed
ruler, in whose honor the temple in Eshnunna was to b built during his
lifetime, had demanded orientation of the temple toward Ur? ... The Gimilsin
Temple confronts us with what appears to b a case of geographical
orientation toward the capital of the overlord (1940:95-96).
Lesson 1 0
This is a text of Ur-Nammu
'
s son and successor, Shulgi, who ruled from 2094 to
2047 BC. It is inscribed on a weight, in the shape of a sleeping duck.
Sign-list and vocabular
_ Sul-gi Shulgi (PN)
_ an-ub corer (?)
_ da side

limmu2 four
Notes
S
ul-gi In older tansliterations the name was read as Dun-gi. It is almost always read Sul
gi nowadays, although there is really very little evidence to permit a decision one way or the
other. It is usually interpreted as "noble (gir) young man (ul)".
The gi-element had an Ir/-Auslaut, although the standard sign-lists do not record any
value in Igir/. Because of this Ir/-Auslaut, he is occasionally referred to as "Shulgir".
This gi-element may be the same gi-element seen in the GN Ki-en-gi.
an-ub The analysis is not clear. Because this word is occasionally written without the an
sign, some think that the an-sign is the divine determinative, and so it is sometimes
transliterated as:
d
ub. Sollberger, for example, explains the word as: "part of the word
(as an emanation of the divine, hence the (divine) classifer)". However, these omissions of
the an-sign occur only in relatively late texts. For simplicity sake, it is transliterated here as:
an-ub, and tanscribed as: anub.
da The cuneiform sign is a picture of the head-upper-shoulder-arm. Its meaning was
extended to mean "side". It is equated in lexical texts with the Akkadian idu, glossed by the
CAD as: "1 . d, 2. side, edge, border, . . . 7. strength".
limmu
2
Numbers can be expressed in two ways in Sumerian: either by a numeral, or by
spelling out the number, using a mixture of logographic and syllabic signs. The word for
"four" in Sumerian was pronounced Ilimmul (or perhaps llimu/, or llima/). This number
was usually expressed by its numeral: . This numeral can be transliterated as limmu.
However, when Sumerian numerals are used stictly for counting, they are usually
transliterated by Arabic numerals, e.g., "4". In certain contexts, when not used strictly for
counting, "four" is expressed by limmu2
'
as in Text 10.
1 09
1 1 0 Lesson 10
Some read the sign as limu
2
instead of limmu
2
. In older transliterations, it is fre
quently transliterated as tabtab. This is still preferred by some modern-day Sumerologists.
1 :
2:
3:
4:
5:
6:
Transliteration
d
Nin-gal
nin--ni
d
S
ul-gi
nitab-kalag-g
I I U
ki
Jg - nm
S
-ma
lugal-an-ub-da-limmu
2
-ba
Text 1 0

T
"
| ! g

'

'
Transcription Translation
Ningal For Ningal,
nin. ani. (r) his lady -
S
ulgi Shulgi,
nitabkalaga the mighty man,
lugal. Urim.a(k) the king of Ur,
lugal.anub.da.(k) limmu.bi.a(k) the king of the four
quarters.
Lesson 10 HI
Comenta
1-2. The text begins with a benefactive to a goddess, even though there is no votive object.
The use of the benefactive is due to the formulaic nature of these inscriptions, coupled with
the idea that all offcial activity, including the regulation of weights and measures, was
conducted ad maiorem deorum gloriam.
3. At some point in the middle of his reign, Shulgi's name begins to appear with the
determinative normally reserved for divine names, *. This (and other) evidence indicates
that Shulgi was deifed, both during and after his life-time; this is discussed further below.
6. There are two genitive constructions in this line, one of which has not been seen
previously: the "anticipatory genitive". Up to this point, to express "the temple of the
king" in Sumerian, one would expect e.lugal.a(k), written -hgal-la. However, Sumerian
has an alterative genitive construction, which puts the noun with the genitive marker frst,
followed by the second noun with a possessive-suffx. A literal tanslation of this
constuction would be: "of the king, his temple". This would be: lugal.a(k) e.ani, written
either lugal-la -i-ni or lugl-la -ni.
This particular expression loosely translates as "the four quarters". A literal translation
would be: "of the corer-and-side, its four": anub.da.(k) limmu.bi. As in the expression
"king of Sumer and Akkad", there is no conjunction between "corner" and "side". Since
these terms are inanimate, the form . bi is used, not . ani.
However, this entire expression is itself the second element of a regular genitive
constuction: "king (of the four quarters)". The frst element in this genitive phrase is
hal. The second element of this genitive phrase is the entire phrase: anub.da. (k)
limmu.bi. The second element is then followed by the genitive marker . ak. The lil of the
possessive-sufx .bi contracts into the tal of . ak, producing Ibak/, and the Ik/, as is the
normal practice, is not written. Thus, a literal translation of this entire expression would be:
"king of [of the corner-and-side, its four] ": [lugal]. [anub.da. (k) limmu.bi] . a(k), producing
"king of the four quarters".
The anticipatory genitive tends to occur in certain fxed expressions (such as in line 6).
In theory, it can be used anywhere a regular genitive could be used, but in practice it is less
common. Since the expression "king of the four quarters" is quite frequent, it is not a
problem to recognize it in context. However, non-idiomatic uses of the anticipatory
genitive can be quite difcult to recognize. The two clues for its presence are: an otherwise
unexplained la/-vowel, followed a little later by an otherwise unexpected possessive
suffx. Several instances of the anticipatory genitive occur in the following texts.
Discussion: numbers
One of the lexical texts found at Ebla is a small tablet giving the names of the Sumerian
numerals from
,
one to ten, spelled more-or-less syllabically. This tablet (TM.75.G. 2198)
was apparently some kind of school or practice text. For "four", the tablet says: l-mu,
presumably for Ilimmu/.
-Typology
It is more common for S-O-V languages to have a genitive construction of the type
I12 Lcsi: !
possessorpossessed then of the type possessed-possessor. The genitive in Turkish, for
example, (an S-O-V. agglutinative language) is of the type possessor-possessed. Sumeri3n
is rather unusual in that the most common pattern is possessed-possessor ("the palace of
the king"), The anticipatory genitive. however, is of the type possessor-possessed ("of the
king, his palace"), and it is possible that it represents the older construction, which perhaps
was in the process of becoming limited to certain stock expressions.
-Weights
Text 1 is a standard inscription, as was Text 7a1b. However, it is inscribed on a
weight. This particular weight does not bear any indication of its value. Text 21a, on the
other hand, is a weight inscribed with its value: mana glna, that is, "5 standard minas".
Weight inscriptions, such as this one, were typically carved of stone in the shape of a
sleeping duck. It is not clear why such a shape was used. The following illustation is of a
duckweight from the NeoAssyrian period.
I
Although stone was the usual material used for making weights, in Assyria bronze
weights were sometimes used instead.
Lesson 10 1 13
- History
The frst Mesopotamian ruler to use the divine determinative before his name was
Naram-Sin, the fourth ruler of the Dynasty of Akkad, who ruled approximately 2254-221 8
BC. Gadd says that "no doubt the vast accession of power and width of sway won by
such a mighty fgure as Naram-Sin helped to make him appear superhuman" ( 197 1 : 619).
The determinatve was used by all the Ur III rulers except Ur-Nammu. It was used only
sporadically by following rulers.
Occasionally, epithets in the royal inscriptions use the word "god". In Text 17a, for
example, Amar-Sin refers to himself as digIT-zid
d
Utu-kalam-ma-na, "the effective god, the
sun-god of his land".
There is also a certain amount of literary material which indicates that the Ur III kings
were considered, or considered themselves, "deifed". However, it is not really known
what this means. The use of the English word "deifed" is rather facile; it is very difcult to
say what this meant to the Ur III rulers or to their subjects. (Gadd says that "vainglory and
popular superstition supported [this policy]" [ 1971 : 619]. ) However, there is a certain
amount of evidence fom royal tombs of the Ur III period to indicate that offerings were
made to the dead Ur III kings, implying that they were worshipped as gods after their
death.
Moorey says that
The most common evidence for the worship of the deceased kings of the Ur
III Dynasty is provided by economic documents describing deliveries to a
place called ki-f-Mg, where liquid offeri ngs to the dead were libated. No
thing specifc is known of these mortuary shrines (1984: 17).
This topic of the deifed king has been discussed by J acobsen:
The deifed king is not a "god" generally; he has the specifc relation to the
country that a personal god has to his ward . . . . The king, as leader of the
country and originator of policy, is the "personal god" of his realm. The
deifcation of rulers in Mesopotamia is accordingly to be understood not in
terms of the qualitative contrast human:divine, mortal:immortal, etc. , but in
terms of function of the king, he is the "genius" of the country (1957: 395
n. 108).
Although the divine determinative is never used before Ur-Nammu's name in any of
his royal inscriptions, it is so used in the Prologue to the Law Code usually ascribed to him.
But this is probably the work of Shulgi, and in any case the Prologue is a rather late copy,
dating fom the Old Babylonian period. Also in the Prologue, he is referred to as the son of
the goddess Ninsun; he is elsewhere referred to in the same way.
-Titulature
It was mentioned above that Naram-Sin of Akkad was the frst Mesopotamian ruler to
use the divine determinative. He was also the first to use the title "king of the four
quaters", in both an Akkadian form and a Sumerian one. The title was not apparently used
by other Akkadian kings. It was used once by a Gutian king, and once by Utu-Hengal of
Uruk, who was overthrown by Ur-Nammu. Ur-Nammu himself did not use the title,
1 14 Lesson 10
presumably because of the limited size of his realm; note also that he did not use the divine
determinative. The title was used by all the other Ur III kings, and afterwards by various
later kings, in an Akkadian or Sumerian form.
- History
Ur-Nammu was killed on the battlefeld, but no specifcs of his death are known; the
literary work entitled "The Death of Ur-Nammu" is terse and unsure at this spot. Woolley
believed that Ur-Nammu and the other Ur III rulers (except the last) were buried at Ur, in a
building complex he referred to as the "Mausolea" of the Ur III rulers. Moorey has
recently questioned this:
The balance of available information, archaeological and textual . . . suggests
that if the kings (and queen-mothers) of the IIIrd Dynasty of Ur were buried
in that city it was not in Woolley's "Mausolea" . . . . Ur is not the only potential
site for these graves, for they might have been in a palace at Uruk, home of
the dynasty, or, less probably, even perhaps at Nippur ( 1 984: 1 8).
It was under the rule of Ur-Nammu's son and successor Shulgi that the Ur III empire
reached its greatest extent; Piot Steinkeller calls him "the true builder of the Ur III state"
( 1 987b:20). There was a great deal of royal building; there was a reform in the calendar,
and much bureaucratic reorganization; Gadd says that "visibly under the impulse of the
king himself a most meticulous system of bookkeeping was instituted" ( 1 971 : 617). Daniel
Snell also points to "a general economic stability during the middle years of the Ur III state,
a stability that seems a likely corollary of the middle kings' l argely successful attempts to
maintain the empire their predecessors had bequeathed them" ( 1 982: 191).
Steinkeller ( 1987b:20-21 ) lists the following among the reforms of Shulgi:
1) the deifcation of
S
hulgi
2) the creation of a standing army
3) the reorganization of the system of temple households
4) the creation of a unifed administrative system for souther and northern
Babylonia
5) the introduction of the bala taxation system, coupled with the creation of
a chain of redistribution centers . . . which served to collect, to process,
and to distribute the state revenues
6) the creation of an enormous bureaucratic apparatus, as well as of a
system of scribal schools that provided highly uniform scribal and ad
ministrative training for the prospective members ofthe bureaucracy
7) the radical reform of the writing system
8) the introduction of new accounting and recording procedures and of new
types of archival records
9) the reorganization of the system of weights and measures
1 0) the introduction of a new calendar, the so-cal led Reichskalender, which
became the offcial caldendar throughout the Ur III state.
It was about half-way through Shulgi's rule when he began to conduct many military
raids. A number of these were directed towards the East, modern-day Iran. The details of
Lesson 10 1 15
these campaigns are rarely known to us, and in fact it is surprising how little historical
inforation we have about Shulgi, especially considering the fact that he ruled for almost
half a century.
Shulgi was also the subject of some thirty hymns, preserved to varying degrees. In
"Hymn B", he boasts: "I leaed the art of the scribe from the tablets of Sumer and
Akkad"; he also refers to himself as "the scribe of Nisaba", the goddess of wisdom and
writing.
1 16 Lesson 10
Text l Oa
supplementary
On a brick from the centre of the Ruins
Notes
This autograph is taken fom Volume 1 of Sir Henry Rawlinson's The Cuneiform
Inscriptions of Western Asia, published in 1 861 . This volume was one of the earliest
collections of cuneiform inscriptions published. At that time, very little was known about
Sumerian. In fact, it seems that it was not until the year 1 869 that the word "Sumerian"
was attached to this language, by Jules Oppert.
It is very difcult to correctly copy texts written in a language which one does not
understand. This means that such autographs are sometimes slightly "off'. In this
particula autograph, some of the signs seem to differ from those in the other texts.
7. en-Unug
kC
!: Funny looking Unug-sign!
The shape of the ki-determinative in this line is quite different from the shape of the ki
sign in lines 9 and 1 1 . It is hard to say how much of this variation is due to the original, or
how much is due to Rawlinson.
8. There is no expected -ma at the end. This is most unusual, and one suspects that this
is an error of Rawlinson, rather than an error of the original scribe.
Lesson 1 1
This text was inscribed on a headdress or wig of diorite. D.J. Wiseman has pointed
out that this headdress was "scored on the underside to ft the rounded head of a statue".
He concludes that it was "intended to be ftted to an actual statue rather than be used
independently as an ex-voto object" (1 960: 168 n.25).
Sign-list ad vocabular
_ Lamar Lamar (DN, fer)
_ Nin-gir-su Ningirsu (DN, masc)
Nane Nanshe (DN, fer)

H
Ba- Baba (DN, fer)
;
+ +
d
Ba--nin-am Babaninam (PN)
| ^
Ur-
d
Nin-gir-su Ur-Ningirsu (PN)
gl zabar bronze
gl_ zabar-dab
S
(kind of ofcial)
bi-li
attaction; headdress, wig

munus woman

dab
s
to hold

di m to fashion, form

y, ba
6
n *
am
Notes
Lamar The reading of the name in the frst line is unclear, since the text is partially broken.
There is a commonly attested goddess whose name is usually written with the kal-sign (Le.,
the kalag-sign); it is variously read by Sumerologists as Lama
2
(or Lama), Lamma
2
(or
Lamma), Lamar, or Lammar. Since this word was apparntly borrowed into Akkadian as
1 1 7
1 18 Lesson 1 1
lamassu, there was probably some kind of Ir/-Auslaut; Lieberman reconstructs the original
form as: ILama/. (The word is further discussed below.)
Virtually everyone who has studied this text has read the frst line as
d
Lamar (re
gardless of the precise transliteration). However, there seems to be present a stoke of
another kal-sign, immediately after the divine determinative:
d
[Kall-kal. This may be a
different writing for the same goddess Lamar, or it may represent an altogether different
deity (a god
d
Kal-Kal is elsewhere attested, but seems to be masculine). For simplicity
sake, and since only a single stroke remains of the problematic sign, the line will be
transliterated as
d
Lamar.
Nin-gir-su Etymologically, "Lord of Girsu". nin is used here in the sense of en, "lord",
as in Text 1 .
Girsu was the sacred quarter of the city and state of Lagash (further discussed in
Lesson 22); it actually lay some distance outside the city prper. Ningirsu was the tutelary
divinity of the state of Lagash. His most famous temple was the E-ninnu (of uncertain
meaning, "House 50" ?).
Ningirsu seems to have been the local name for the god elsewhere worshipped as Nin
urta, a god originally of agriculture and storms, but also of war. The two were probably
independent deities who were very early identifed with each other.
Nane This name is read by some Sumerologists as Nazi. She was a daughter of Enki,
and the goddess of Lagash. She was consulted for the interpretation of dreams. When
Gudea, the ruler of Lagash, had an odd dream in which a mysterious fgure appeared, it
was Nanshe that he turned to for the explanation of the dream.
The cuneiform sign representing her name is the sign for the city of Sirara (one of the
places where she was especially worshipped), with an inscribed -sign. means
"fsh"; the sign is in origin the picture of a fsh. This and other evidence indicates that
Nanshe may originally have been some kind of fsh-goddess.
Ba-ba6 The reading of the last sign is uncertain. The name is variously transliterated as:
Ba-y, Ba-, Ba-bUl l (and Ba-bul2 )
'
and Ba-w,. She was the wife of Ningirsu, and
hence the city-goddess of Lagash. At times, Inanna herself is referred to as "Ba-".
zabar In older transliterations, each of the three individuals signs forming this compound
logogram is separately transliterated: ud-ka-bar. The etymology and writing are discussed
below.
zabar-dab
S
The etymology is discussed below. The function of this offcial is not too
clear. Jacobsen refers to him as "the ofcial in charge of the bronze (table-wares, cups,
knives etc. of a large establishment, and possibly of the bronze weapons as well)"
( 1 957 : 382 n. 55).
The word was borrowed into Akkadian as zabardabbu. The CAD simply tanslates
the Akkadian term as: "an offcial". After a long discussion, it concludes with the remark
that this offcial was "(possibly), originally the weapon carrier of the king".
Lesson 1 1 1 19
bi-li The basic meaning of bi-
g
appears to be something like "chan" or "attraction". (The
Akkadian equivalent, kuzbu. is glossed by the CAD as: "luxuriousness, abundance,
attractiveness, charm, sexual vigor".) It also has the derived meaning "headdress" or
"wig". The gudug-priests discussed in Lesson 19 are attested as wearing a bi
-
g
.
munus Also transliterated as m. Particularly in older works, it is tansliterated as sal.
dab
S
In older tnsliterations, di b (and, incorrectly, dib
4
).
dm While du is used for the (re)constuction of more solid objects, such as palaces,
temples, etc. , dim is nonally used of smaller, hand-made objects.
The Akkadian equivalents for du and di m are not neat. di m is nonally equated with
banu, but du is equated both with banu and epeu.
'
Am This sign is composed of two elements: g which nonally has a syllabic reading ,
and _ g which nonally has a syllabic reading an. It is not clear how these two signs
came to represent (together) the value I amI.
120 Lesson 11
Text 1 1
-

_ . . .

=
Transliteration
1 :
d
Lamar
2: nin-.-ni
3: nam-ti l
4:
d
S
ul-gi
5: nitab-kalag-g
6: lugil-Urim
s
ki
-ma-ka-e
7
d
B b " : --
6
nm-am
8: zabar-dab
S
9 U
d
N'
-
, : J- -- -gr-su
1 0: en-ki-.

_
d
Nane-ka-ke
4
1 1 : bi-l-nam-munus-ka-ni
12: mu-na-dim
Comenta
Lesson 1 1
Transcription
Lamar
nin. anL(r)
nam.til
S
ulgi
nitabkalaga
lugal.Urim.ak.a(k).e
Babaninam
zabardab
Umingirsu
Translation
For Lamar,
his lady -
for the sake of the life
of Shulgi,
the mighty man,
the king of Ur
Babaninam,
the zabardab
of Ur-Ningirsu,
121
en.kLaga.a.N ane.k.ak.e the beloved lord of N anshe -
biILnam.munus.(a)k.anL0 her "beauty of womanhood" -
mu.na.(n.)dim.0 he fashioned.
3-6. The essence of this line is: "for the life of Shulgi". This would be expressed by:
[nam.til.
S
ulgi.k] .e. However, this nominal phrase is complicated by the presence of two
appositives. The frst is the noun-adjective combination, nitab kalag-gf. The second is the
genitive phrase "king of Ur". By itself, this last geni tive phrase would be expressed by:
lugal.Urim.ak. This nominal phrase may be diagrammed as:
[nam.til] [
S
UIgi
J
.ak.e
nitabkalaga
lugal.Urim. ak
This results in a succession of two genitive markers, followed by the marker for the
teninative case: . ak.ak.M. In the script, this is refl ected as: . . . Urim
s
kt
ma-ka-e.
Since the Ikl of the frst genitive marker .ak is followed by a vowel, it is pronounced
and written in the script. The Ikl of the second genitive marker .ak, however, is syllable
fnal before a consonant. In such cases, the Ikl does not show up in the script. The
problem of the amissable word-fnal consonants was discussed earlier. To some degree,
that same problem is present when these consonants are syllable-fnal, not just word-fnal.
That it, it is not sure if these consonants were pronounced or not: it is not known if this is
an orthographic or phonological problem. The sequence .ak.ak.e, for instance, is common,
but the Ikl of the second genitive marker does not appear to ever be written in any of these
Occurrences. In general, it seems that the Ikl of the genitive marker does not appear in the
script, when it is in syllable-fnal position followed by a consonant. It is transcribed here
within parentheses.
The ka-sign includes the Ikl of the fi rst genitive marker, and the lal of the second
genitive marker. This is yet another instance where the script does not follow the
morphology, but rather approaches the syllabic structure of the spoken language. This
122 Lesson 1 1
sequence may well have been syllabifed something like /u-ri-ma-ka(k)-se/.
d
Ba-Ba
o
nin-am is a personal name, "(the goddess) Baba is a lady", or perhaps "Baba,
is queen". The divine determinative "goes with" the name of the goddess Ba-ba
6
'
not with
the personal name itself.
Sumerian has two ways to express the copula. The frst way, which is in fact
relatively uncommon, is to infect the verbal root meaning "to be" (me); it thus behaves like
a regular verb. This is sometimes called the "independen t copula". The second way, which
is much more common, is to use a reduced form of thi s root as a sufx, instead of as an
independent verb. This is called the "enclitic copula". For the third person, this consists of
.am (usually wrtten |/ am) suffxed to the second element of an equational sentence.
Thus, "Nanna is king" is: Nanna.lugal.am, written:
d
Nanna-Iugal-am; "Baba is queen" is:
Baba.nin. am, written,
d
Ba-ba
6
-nin-am. (In pre-Ur III texts, the enclitic copula is regularly
written
;
,read as am
6
'
)
To judge from previous writings of morphemes beginning with a vowel, one might
have expected a writing something like
d
Ba-l-nin-nam, or
d
Ba- -nin-na-am. How
ever, this is a case where the script is morpheme-bound; the one sign am regularly
expresses the morpheme . am, and there is no graphic reduplication of the preceding
consonant. (Nor does the writing refect the presumed syllabic stucture.)
In general, names of men in Sumerian are construed with the names of gods, and
names of women are construed with the names of goddesses. However, there are
exceptions to this rule ("Ur-Nammu", for example, is construed with the name of the
goddess Nammu), and there seems to be no evidence that the zabardab-offcial was ever a
woman. Thus, it cannot be determined whether Babaninam was a man or woman.
8. The word zabar-dab
S
is composed of two elements, zabar, "bronze", and dab
S'
a verbal
root meaning "to grasp". dab
S
here is probably an active participle. In general, verbal roots
in Sumerian have two participles: an active participle in .0, and a passive participle in .a
(the same nominalizer seen previously). The use of these rather conventional terms is not
without problem, but in general the active participle denotes the doer of the action. Thus,
dab
S
is "the one who grasps", "he who grasps". The passive participle denotes the result of
the action, or the one acted upon. From sar, "to write", the passive participle sar-ra means
"something written".
Some Sumerologists refer to adjectives in .0 (such as man) as active participles, and
adjectives in . a (such as kalag-gf) as passive participles. It is not sure if this is a valid use
of these terms; part of the problem is the diffculty in defning the different categories of
root in Sumerian: verbal, nominal, etc.
As do participles in English, the participle in Sumeri an can take a direct object. In this
case, zabar is the direct object of dabs
'
Thus, an etymological translation of zabar-dab
S
would be: "the one who grasps the bronze", or "he who grasps the bronze" (the object
precedes the participle, just as the direct object (patient) precedes a verb). However, this
particular phrase may have been felt as one unit, since it was borrowed into Akkadian as
one word, zabardabbu.
9. The PN Ur-Ningirsu means "the man of Ningirsu". The DN Ningirsu itself means
"lord of Girsu". Therefore, the name might be understood as: [Ur] . [Nin.girsu.k].ak. The
Lesson 11 123
frst genitive marker is for [Nin] . [
G
irsu] .k; the second is for [Ur] . [Ningirsu.k] . ak.
However, as discussed in Lesson 1 concering the name Ur-Nammu, there is some
indication that the genitive marker in PNs was lost, and so it is not indicated in
tanscription.
Lines 7-10 form the ergative nominal phrase, expressing the agent of the tansitive
verb. The ergatve case-marker .e appears at the end of line 10. The nominal phrase is
complicated by the presence of all the appositives. Line 7 is a PN, Babaninam. Lines 8-9
re an appositive, describing Babaninam as the "zabardab of Ur-Ningirsu". This would be
expressed as: [zabardab] . [Uringirsu] .k, but Ur-Ningirsu himself is described as "the
beloved lord of Nanshe", an, .. appositive. "Beloved lord of Nanshe" is: [en.ki.aga.a] .
[Nane] .k, a more complicated genitve phrase than any seen up to now: ki.aga.a modifes
en, and the combinaton is itself the fst element of the genitive phrase. Thus, the nominal
chain describing Ur-Ningirsu is: Uringirsu, en.ki.aga.a.Nane.k. All of this is the second
element of a genitive phrase, with zabadab being the frst element: [zabardab] . [Urningirsu,
en.ki.aga.a.Nane.k] .ak, all of which is an appositive to Babaninam. This may be dia-
gr

b
am
Uringirsu . ak ]
.e

en.ki.aga.a.Nane.k
The writing of the end of the nominal phrase is as expected: . . .
d
Nane-ka-e
4
.
1 1 . munus is "woman", nam-munus is an abstact, "womanhood".
This line must mean something like "her beauty of womanhood", that is, "her woman's
beauty". If so, then it is a genitive phrase, followed by a possessive-suffx: [bili.nam.
munus.ak] .ani.
The basic rule for the genitive marker as presented up to now has been: /ak/ after
consonants, /k/ after vowels. Here, however, there occurs: nam-munus-ka-ni; the vowel
/a/ of the genitive marker /ak/ does not seem to appear in the writing.
Such writings -where the /a/ of the genitive maker does not appea after a consonant
- are not uncommon; the next example is in a formulaic phrase appearing in Text 13. It is
diffcult to say whether such writings tell us something about Sumerian orthography, or
Sumerian phonology, or Sumerian morphology; several interpretations ae possible. In
order to make the written form ft our understanding of the grammar, one school of thought
would read the frst sign as munusa, instead of munus. This would produce munusa-ka-ni,
accurately refecting munus.ak.ani. Paallel phenomena occur outside of the Ur III corpus.
For example, in the inscriptions of Gudea - inscriptions highly localized to one time and
place - "his king" is expressed by both lugal-ni and lugal-.! -ni. In order to make lugal-ni ft
more accurately our understanding of Sumerian, some scholas would read the two signs as
lugala-ni, and not lugal-ni.
This school of thought was particulaly adumbrated by Adam Falkenstein, who saw
simila phenomena elsewhere in Sumerian grammar. He coined the term "
U
berh!ngende
Vokale" (in English, "overhanging" or "overlapping" vowels) to describe just such
writings. This school would thus see such writings as an orthographic problem. (The
standard sign-lists, however do not seem to recognize a reading *munusa for the sign in
124 Lesson 1 1
question.)
A different view sees this as a morphological (or phonological) problem. Mamoru
Yoshikawa feels that the genitive maker was sometimes Ikl after a consonant, not only
after a vowel. However, he cannot posit rules for the distibution of lakl and Ikl after
consonants. Yoshikawa also believes that there ae cases where the genitive maker is
I ak/, not Ik/, after a vowel.
This is not a problem which can be solved here. But it must always be kept in mind
that the writing system of Sumerian never accurately refected the spoken language. It is
possible that although a scribe spoke /lugalanil, he was perfectly happy to write l! al-ni,
because with just these two cuneiform chaacters, he knew what to read. Why bother to
write an i-sign if the context makes the presence of a spoken lal obvious?
This entire line is the direct object (patient) of the verb dim, and therefore is in the
absolute case.
Discussion: structure
It is diffcult to see the basic structure of this text, because of the presence of so many
appositional phrases. Its substance, however, is: "Babaninam fashioned a wig for Lama,
for the sake of the long life of Shulgi":
[Lamar, nin.ani] . (r)
[nam.til.
S
ulgi, nitab.kalaga,
lugal.Urim.ak.a(k)].e
[Babaninam, zabardab.Umingirsu,
en.ki.aga. a.Nane.k.ak] .e
[bili.nam.mun us. (a )k.ani] . 0
mu.na.(n.)dim.0
-Writing system
benefactive
purpose
agent
patient
verb
The innocuous-looking word zabar illustates some of the inticacies of the Sumerian
writing system. There are no metals native to Sumer; rather, all had to be imported. Thus,
zabar i s not a native Sumerian word; it was borrowed fom some unknown language.
Hallo says "in general, it may be supposed that the basic metal names ae non-Sumeran
'Kulturworter' or 'Wanderworter' which were adopted together with their referents"
( 1963: 140). In Akkadian, the word for bronze is siparru. Akkadian may have borrowed
this word independently from the same language that Sumerian borrowed it from, or, much
more likely, borrowed it directly from Sumerian. In either case, the form siparu is a little
odd; it would seem to derive from */sipa/, not Izaba/.
One way to solve this discrepancy is to assume that in ealier Sumerian the word for
"bronze" was, in fact, Isipar/, and that Izabarl rpresents an inner-Sumerian development.
The change of Ipl Ibl is not surprising; voicing of inter-vocalic voiceless consonants
happens in many languages. The change of initial Isl IzI is less easily explained, but
there ae other paallels to this change in Sumeran. The difference in vocalization between
the two forms is more interesting. As will be discussed i Lesson 20, there is a fair amount
Lesson 11 125
of evidence to show that Sumerian has undergone a rather wide process of vocalic
assimilation. In words originally containing two vowels of differing quality, one vowel has
assimilated to the quality of the other. In our particular case, an original */i-al has become
la-al.
Thus, Izabarl can be derived fom Isipar/, using sound-changes which are elsewhere
attested in Sumerian. (Unfortunately, not enough is known to date these sound changes.)
This then would represent a case where the Akkadian word has actually preserved a more
archaic form of the word than has Sumerian. Pesumably Akkadian borrowed it from
Sumerian before these changes took place. Similar instances will be discussed in Lesson
20.
The pronunciation of this word in historic Sumerian as Izabar/, and in Akkadian as
Isiparru/, is known fom lexical lists, where these words are spelled out as za-bar and si
par-ru. A next question is, how does the prnunciation Izabarl "derive" from the three
signs ud-ka-bar? In the word nidba, for instance, there was no obvious way to phonetically
relate the prnunciations Inindabal or Inidbal to any pronunciation of the individual signs;
that is, the word was more than the sum of its parts.
Since one of the three signs forming the word for "bronze" is the bar-sign, it seems
reasonable to assume that Izabarl derives phonetically from these three signs. But how?
One possibility might be to read ud-ka as za
x
. This type of approach is favored by many
Sumerologists, who t to make the writing system better ft Sumerian pronunciation.
However, there does not seem to be any other, independent, evidence which would justify
positing a reading za
x
for this sign, and the standard sign-lists do not recognize such a
value.
However, a further complication must be intouced. In the earliest Sumerian, the
word for "bronze" is not, in fact, written ud-ka-bar. Rather, it is regularly written as
KAxUD-bar, that is, with a ka-sign containing an inscribed ud-sign, followed by the bar
sign. For example, in a ryal inscription of king Uruinimgina of Lagash, the word for
bronze appears as: L.
It is difcult to explain such a writing. Some scholars have posited a reading for
KAxUD; this is accepted by some sign-lists, although with reservation. However, Fal
ken stein has pointed out evidence that KAxUD can be read as sil
9
. This reading is
accepted by the standard sign-lists. Perhaps, then, the word should be transliterated as:
si
19
-bar. Armas Salonen, in fact, tansliterates this word as si
l9
-bar for the older period,
but as zabar for the "nachsumerische" period ( 1 961 : 108). I this interpretation, which is
probably corect, si
l9
-bar represents an older pronunciation of the word. But the same
cuneiform signs continued to be used, when in spoken Sumerian the word had changed to
Izabar/.
-Writing system
The taditional interpretation of the nin-sign ( ) is that it represents the sign for
"woman" (munus, ), followed by the sign for "clothing" (tug, .That is, the sign is
a woman wearing a (special) kind of clothing, to mark her elevated status. This
interpretaton has been questioned by Robert Biggs, who says that in the earliest Sumerian
texts, the tug-sign is different from the component foring the second half of the nin-sign;
12
6 Lesson 11
this is tue to some degree even i n Old Babylonian texts. This leaves the origin of the nin
sign up in the air.
-Overhanging vowels
The problem of the overhanging vowels has generated a go deal of polemics in the
course of Sumerological studies (the latest discussion is Yoshikawa 1980). This over
hanging vowel seems to b mostly la/, although individual cases of overhanging le/, liI,
and lul have been posited. Probably most Sumerologists accept their existence, but there
are dissenting opinions.
In Falkenstein's view, the readings in lal represent older forms of Sumerian words.
That is, at one time these words were pronounced with a fnal la/. The word for "king",
for example, was originally *Ilugala/. At some (prehistoric) point, these fnal lals were
dropped; the word for "king" became Ilugal/, but the sign used to represent this word
could be used for either the newer value Ilugal/ or for the older value *Ilugala/. Since
both values co-occurred, a scribe could write "his king" as either lgal--ni, or lala-ni,
both representng /lugalanil. (Presumably, the prehistoric form would have been *Ilugala
anil or *Ilugala-nil. )
Other Sumerologists question their existence; they see varying phonetic factors at
work. It was mentioned above, for example, that Yoshikawa believes that the genitive in
Ikl occasionally occurred after consonants, not just vowels; thus, "the son of the king"
might have been pronounced Idumulugalk/. However, Yoshikawa could not state any
general rules for the distribution of Ikl and lakl after consonants. For other overhanging
vowels (e.g., in the verbal system), he has other explanations.
In writings of the lgal-ni type, it has been posited that the lal was dropped:
*Ilugalanil ) Ilugalnil (this was Poebel's position). Without going into details, it can b
seen that such an explanation raises more questions than it answers, such as the co
occurrence of writings like lugil-ni and lugil-f-ni in the same time and place.
Is there any independent evidence which justifes the view that certain signs contain an
overhanging vowel? Here the evidence varies, and is difcult to interpret. The fact that the
nominalized forms of the verb "to love" are written both ki-5g5
2
and ki-g-g might seem to
indicate that this one sign can be read as lagl or laga/. However, it is also possible that
g is here a phonetic complement, and the form should b understood as: ki-5g

g
a
; or it
might be that these writings refect phonological problems of particular roots ending in a
vowel.
The Akkadian lexical tradition is likewise ambiguous. Lexical lists do provide
readings with lal for some signs, but for the most part they do not (for example, they show
no evidence of a reading lgala). And, some of these readings with an overhanging vowel
may very well result from the Akkadian scribes encountering the same problems in the
writing system that we feel. These scribes may have anticipated some modem
Sumerologists, by generating readings in la/, in order to make the writing system more
closely ft the pronunciation.
Part of the problem may result from a misunderstanding of the nature of the Sumerian
writing system. Because the expression "malt house" is written -bappir, for example, and
Lesson 1 1 127
not written -bap ir-ra (for the assumed e.bappir.a(k , Falkenstein would say that the
second sign should be read as bap ira. But it is easier to say that the Sumerian scribe felt
no need to write any indication of the genitive marker; such scribes were content to write
bap ir, even if they pronounced it /ebappira(k)/. Falkenstein's school is an attempt to
make the writing system more closely resemble a transcription of speech, and this is not
how the writing system should be understood.
It is tue, however, that there are other problems to be resolved. For example, "in the
land of Sumer" is normally written kalam-ma in the Gudea texts, for kalam.a. But once,
apparently, this locative phrase is written: kalam. Falkenstein would read this kalama.
Similarly, "on the tablet", written just dub, would be read by Falkenstein as duba. These
writings raise questions, but they may simply refect an earlier period in Sumerian
orthography, when it was not necessary in general to write case-endings.
- Loan words
As just discussed, "bronze" is zabar in Sumerian, siparru in Akkadian. The word for
"copper" is urudu in Sumerian, weru in Akkadian; they are usually spelled out in lexical
lists as !-ru-du and -ru-.. The ultimate origin of the word(s) is unknown. Both urudu
and wen may refect one pre-Sumerian substrate word; they have even been connected
with the Indo-European word which ultimately appears in English as the adjective "red".
Curiously, in late Akkadian the word for "copper" also appears as urudG. Eduard Kutscher
says that "this 'Akkadian' word was artifcially coined by Sennacherib's scribes (and used
only by them) from Sumerian urudu ( Akkadian wen1) 'copper'" ( 1982:225).
- History
Lamar is well-attested as an intermediary or intercessory goddess. She appears on Ur
III and Old Babylonian seals, introducing a worshipper to a higher god or goddess.
Because of this function, the name Lamar becomes almost a generic word for "protection".
Thus, there occur personal names of the type: Lugal-
d
Lamar-
RlI O
, the king is my pr
tection". Borrowed into Akkadi an, the word lamassu is glossed by the CAD as: "prtec
tive spirit". Von Soden, however, has questioned the traditional derivation oflamassu from
Lamar.
- History
Other wigs have been found at various sites in Mesopotamia, although none of them
bear an inscription. For example, a wig of steatite was found at Uruk. Only 2.5 cm. long,
it was apparently designed to ft a statue (the dating is uncertain; before Nabonidus). More
recently, somewhat similar wigs have been found at Ebla.
I7 1sOn!!
Text I I a
supplementa
Lesson 1 2
This text was inscribed on the foot of a vase of marble. No photograph is available.
Sign-list and vocabulary
_d _
d
Bi I-g-me Gilgamesh (DN, masc) ( bi !)
_ En-dim-gi Endimgig (GN ?)
ud U) day
+|
__ bi
ib

b
da
Notes
mu-sar-ra inscription
u . . . ur to erase
nam . . . kur
s
<
s
) to curse
d
BU-ga-me This is now the preferred reading (or at least the more original reading) of
the name more familiarly known as "Gilgamesh". The apparent meaning of the name is:
"the old man (bi1-g is (now) a young man (mM)". It is not known when the change of
initial /b/ ) / g/ took place; an Old Babylonian omen text has the spelling
d
Ge-el-g.
Some read the last sign of the name as mes.
mu-sa-ra sar is "to write". sar-ra sar.a, the passive participle: "something written".
mu has many meanings (e. g., "name"); mu. sar.a. is something like "a written text", or
"inscription". It was borrowed into Akkadian as musarU, tanslated by the CAD as: "I .
object bearing an inscription, 2. inscription".
u . . . ur This is a compound verb. u means "hand", and ur means something like: "to
move or drag (something)". u is thus the historical patient of ur: "to move the hand over",
i.e., "to erase". The verb takes its complement in the locative case. (The Akkadian
1 29
130 Lesson 12
equivalent is pasatg, glossed by AHw as: "tilgen, ausloschen".) ur also enters into the
formation of other compounds, for example: gis . . . ur, "to harrow" (literally, "to drag wood
(over the ground)".
Some believe that the root ended in lul, and so it i s also transliterated as su . . . uru
12'
nam ... kur
S
nam is apparently the (historic) direct object (patient) of kur
5
; the meaning is
something like "to cut a decision (against)". It usually takes its complement in the
comitative case. The word is further discussed below.
Lesson 12
131
Text 1 2
132
Transliteration
1 :
d
Bi I-g-mM
2: En-di m-gig
ki
3: lugl --ni
4: Ur_
d
Nammu
5: nitab-kalag-
6 I I U '
ki
: -!g - nm
S
-ma
7: lugal-Ki-en-gi-Ki-uri-ke
4
8: ud _
d
Nanna
9: mu-du-
10: nam-til-Ia-ni-e
1 1 : -mu-na-ru
12: lu mu-sar-ra-ba
1 3: u-bi -i b-ur-
14:
d
BiI-g-me-
1 5: nam-b-ba-da-kur
)

Comenta
Lesson 12
Transcription
Bilgame
Endimgig.(ak?)
lugal.ani. (r)
Umammu
nitabkalaga
lugal. Urim. a(k)
lugal. Kiengi. Ki uri.k.e
ud e.Nanna.(k).0
mu.(n.)du.0.a.a
nam. til.anUe
a.mu.na.(n.)ru
lu musara. bi.a
u.bi .b. ur.e.0.a.(d)
Bilgame.e
nam.be. ba.da.kur.e.0
Translation
For Gilgamesh
of En dim gig,
his king -
Ur-Nammu,
the mighty man,
the king of Ur,
the king of Sumer and Akkad -
when he built the temple of
Nanna,
for the sake of his long life
he dedicated a votive offering.
May Gilgamesh curse
the man who erases
this inscription!
2. The meaning of this line is unclear. The frst editor of the text read it as en D
I
M.GIG
ki
,
"lord of D
I
M.GIG", an otherwise unattested place name. However, the editor added the
comment: "I cannot understand Gilgamesh's epithet in line 2".
It has been pointed out that there is an apparent GN, En-di m-gig
ki
, which occurs at
least twice in Sumerian texts. There, the en-sign is apparently an element of the place
name, not the word "lord".
Presumably, lines 1 -2 form a genitive phrase: "Gilgamesh of Endimgig". This is,
curiously, the same construction found in one of the other occurrences of the GN:
d
Nin
ubura EN.D
I
M.GIG
ki
, tanslated by J. van Dijk as: "Ninubur von (?) EN.D
I
M.GIG
ki
".
A parallel to this construction (DN and GN in a genitive phrase) occurs in Text 17:
d
Nanna Kar-zid-da lugal-ki-.M-M.-ni-i, "To Nanna ofKarzida, his beloved king".
In any case, one might have expected a fnal g-sign, to express the lal of the genitive
element.
8. ud introduces a subordinate, temporal clause: "when he built the temple of Nanna".
Sumerian does not have many different kinds of subordinate clause formation. The most
common type is a temporal clause. As is frequently the case in Sumerian, it is fairly easy ,to
recognize the surface form of the construction, but it is a little harder to understand the
grammar behind the written form.
The simplest temporal clause consists of: ( 1 ) a relative marker; (2) a relative clause
nominalized in .a, which stands in apposition to the relative marker; (3) a locative case-
Lesson 12 133
marker in .a.
The most common relative marker is the noun ud (literally, "day"); others also ocur.
Occ asionall y the relative marker is deleted.
The essence of the relative clause is: "He built the temple of Nanna". This
independent sentence would be: e. Nanna.(k).0 mu.(n.).du.0. To form the relative clause,
this independent sentence is nominalized in .a. It is then placed in apposition to the relative
marker ud: ud [e.Nanna.(k).0 mu.(n.)du.0].a.
It is instuctive to compare this relative clause with the relative clause in Text 7. In that
text, "the one who built the temple of Enlil" was expressed as:
lu [e.Enli1. a(k). 0 i.n.du.0].a.
The constuction here is much the same, but the relative marker is ud instead of lu:
ud [e.Nanna. (k). 0 mu.(n.)du.0]. a
Finally, the entire complex is put into the loative case, marked by . a (the second . a in
the transcription of line 9).
The function of the locative in .a here is "on": "on the day that". It is also possible to
fnd other case-markers, such as the ablative case-marker ta ("from, since"), or the
terminative case-marker Se ("until").
A literal tanslation of this entire clause would be: "on the day that he built"; this
captures both the force of the Sumerian loative, and the force of the relative clause. In
more idiomatic English, however, one may say "at the time when", or simply "when":
"when he built the temple of Nanna, he made a votive offering".
In this temporal clause, the verbal chain uses the conjugation-prefx mu, and the
personal-affx .n does not appear in the script. In the relative clause in Text 7, the
conjugation-prefx was 1. and the personal-affx .n appeared in the script. It is not easy to
understand the reasons for such alterations.
Although the transcription indicates both the . a of the nominalizer and the .a of the
locative case-marker, it is reasonable to assume that some kind of voalic contaction tok
place; one never fnds two /als in the script in such a constuction.
Lines 8-9 form the subordinate clause; lines 10- 1 1 form the main clause. In Sumerian,
the subordinate clause regularly precedes the main clause.
12- 15. These lines express a curse; essentially the same wording occurs in other Su
merian inscriptions.
Lines 1 2- 1 3 are a relative clause, serving as the complement of the verb in line 15. Its
meaning is: "the one who erases this inscription". The relative marker is lu; the relative
clause is marked by the at the end of line 1 3.
Su . . . ur is a compound verb. bi is another conjugation-prefx, in addition to the mu and
1 already seen. It is almost always written with the bi-sign. It will be further discussed
below.
All the verb forms seen up to this point have been in the gamn! aspect, used to express
action in the past. The verb form in this line, however, is being used to express a future
value: "whoever will erase", or "whoever might erase". Therefore, it is put in the man
aspect.
It is common to speak of the "gamtQ-root" and the "marf-root". The man-rot is
1 34 Lesson 12
foned fom the bamt!-root in several different ways. The fonation used for any par
ticular verb is lexical; that is, it is not predictable. It is not clear exactly how many classes
of maru fonation exist. Y oshikawa has established three different classes:
( 1 ) "Reduplication" class. The man-root is foned by (graphically) reduplicating the
bamt!-root. Thus, "to retur": bamt!, g4; maru, g-g.
Roots of the patter eve seem always to lose their fnal consonant when reduplicated.
Thus, "to place": bamt!, g; man, .M-.M (always written with the g!-sign).
(2) "Alternation" or "replacement" class. An entirely different root is used for the
maru. This root is non-predictable fom the bamm-root. Thus, "to speak": bamm, .!g
4
;
maru, . dug
4
and are two entrely different signs. However, there are cases where a
bamt!-rot and a maru-root will be written with the same sign. Thus, "to go": bamt!, gin;
maru, duo The gin-sign and the du-sign are the same! In such cases, it is only the gramma
tical context which indicates whether the sign is to be read as the bamt!-root or as the
maru-root.
(3) "Affxation" class. This is foned by addition of .e to the bamt!-root. This .e is
variously referred to as the "maru-element", the "maru-affx", or the "man-suffx". Thus,
"to build": bamt!, du; maru, du-. This class is the most common fonaton of the man.
It is the fonation used with the verb in Text 1 2, ur: ur.e. Here, however, the maru-suffx
has contracted into the nominalizer . a, and so it does not show up in the script.
In general, the particular maru fonation for any specifc verb is not always known.
And, some roots fall into two (or even all three) classes. As mentioned above, ur is a
member of the affxation class, but reduplicated man fons also seem to occur. In later
Sumerian, combinations of these classes sometimes occur; e.g., a reduplicated root
followed by the maru-suffx. These cases have not all been explained.
The use of the maru, instead of the bamt!, entails rather complex changes in Sumerian
morphology, particularly in the distibution of the personal-affxes. For a tansitive verb:
in the bamm, the personal-affx slot before the verbal root cross-references the agent, and
the personal-affx slot after the verbal root cross-references the patient. In the maru,
however, it is just the opposite. The personal-affx slot before the verbal root cross
references the patient, and the personal-affx slot after the verbal root cross-references the
agent. For example, "The king built the house" is:
(1) lugal.
e
e. mu.

.du.
But, "The king will build the house" is:
T
W
(2) lugal.e e.0 Lb.du.e.0

In (2), the ergative case marked in .e is cross-referenced by the .0 after the maru-
suffx. The absolute case marked in .0 is cross-referenced by the .b before the verbal root.
.b is used here to cross-reference inanimate antecedents; .n is used to cross-reference
animate antecedents.
That is, the case-markings on the nominal partcipants in the sentence are the same in
both aspects: lugal.e and e.0. However, the use of the personal-affxes is quite different.
Thus, in the verb fon in this line, u.bi.b.ur.e.0.a.(d), the .e is the maru-suffx, and
the .0 cross-references the third-person agent. (This analysis of .e.0 is not universally
accepted, and a different view will be mentioned below.) The .b. before the root cross-
Lesson 12 135
references the u, which is the (historic) patient of the verb ur.
Finally, the combination of bi . b. is quite frequent, and usually written with the bi -sign
followed by the i b-sign, as in Text 1 2.
The verb form may be summarized as:
u bi . b . ur . e . 0 . a
( 1 ) (2) (3) (4) (5) (6) (7)
( 1 ) nominal element of compound verb
(2) conjugation-prefx
(3) personal-afx cross-referencing patient )
(4) verbal element of compound verb
(5) man-sufx
(6) personal-afx cross-referencing agent (unexpressed)
(7) nominalizer, foring relative clause
mu-sar-ra-ba is for musara.bi.a, "on this inscription" . . bi is a demonstrative, meaning
"this". It is sufxed to its noun. It is identical in for (and probably in origin) with the
third person inanimate possessive-sufx .
. a is the locative case-marker. Imusara.bil is in the locative case because it is the
complement of the verb u . . . ur, which takes its complement in the locative. The Ii/ of Ibi/
has contracted into the lal ofthe locative.
To sum up, lines 1 2- 1 3 are a relative clause, meaning, "the one who erases this
inscription". This entire clause functions as the complement of the verb nam . . . kur
5
in line
1 5.
The compound verb nam . . . kur
5
most frequently takes its complement in the "comi
tative" case. The word comitative comes from the Latin word cum, meaning "with". It is
not easy to defne or even to adequately describe all the uses of the comitative case, but it
often expresses ideas such as "along with": hl-da, "with the man". Besides this use, many
verbs in Sumerian take their complements in the comitative case. This usage is not usually
predictable, and must be listed in the dictionary.
The basic form of the comitative case-marker is .da, written: da.
There is a prblem here, however, because there is no overt marker of the comitative in
the script: one would expect to fnd a da written at the end of line 1 3. One explanation for
its absence is to assume a phonological process similar to that of the dative case-marker .ra.
That is, Idal Idl after a vowel, and word-fnal Idl is not written. However, in the royal
inscriptions of Gudea or Ur Ill, the comitative case-marker normally appears written as da,
even after a vowel; cases where an expected da does not show up in the writing only occur
in earlier Sumerian.
The problem of the presence or absence of the comitative case-marker in this line will
be further discussed below.
The scope of the assumed comitative case includes all of lines 1 2 and 1 3:
[Iu muara.bi.a u.bi.b.ur.e. 0.a] . (d).
1 4. The agent is marked in .e. The writing is morphemic; there is no attempt to graphically
reduplicate the fnal IU of
d
Bi I-gi-mM. Such morphemic writings are especially common
136 Lesson 12
with proper names; the name was felt as a unit, and the grammatical case-endings added
directly to the complete unit.
15. b is a form of the "desiderative" or "precative" modal-prefx. As discussed in Lesson
1 , the very frst element in the Sumerian verbal chain is an optional modal-prefx; there are
several of these.
The desiderative mood expresses wishes and indirect commands in the third person:
"Let him/them, may helthey", etc. The basic form of the desideratve is Ibel (written b);
before the conjugation-prefx ba, it regularly becomes Ibal (written b).
One of the more difcult questions in Sumerian morphology is the nature of the root
used after particular modal-prefxes. Some modal-prefxes use the bamm-root, others use
the man-root, and still others use both, under conditions which are not always clear. A
second problem is the use of the personal-afxes. In some moods, the pre-verbal root slot
cross-references the agent. In others, it cross-references the patient. (This diferentiation is
apparently irrespective of whether the bamt!-root or the man-root is used.)
With transitive verbs, M is regularly construed with the mam-root; with intransitive
verbs, it is construed with the bamm-root. kur
S
is a member of the afxation class, and so
the form is: kur
S
-' The writing is morphemic in this line; in other cases the fnal consonant
ofthe verbal root may be graphically reduplicated (e.g., kur
S
-re).
In the desiderative, the agent is cross-referenced by .0 after the maru-sufx.
As with all compound verbs, the frst element of the compound (here, nam) precedes
the entire verbal chain, including the modal-prefx.
ba is a conjugation-prefx not seen previously. It is usually assumed that it is related to
the conjugation-prefx bi , an example of which occurred in line 1 3. It is further discussed
below.
da is the dimensional-prefx which cross-references the comitative case. Here it crss
references the comitative case marked by the presumed .da at the end of line 1 3. The
original meaning of this compound verb may have been something like "to cut a decision
against".
To summarize the verb phrase, nam . . . kur
S
is a compound verb, with nam the (his
toric) patient. b is a form of the modal-prefx for the desiderative; with transitive roots, it
uses the maru form of the root. ba is a conjugation-prefx. da is a dimensional-prefx,
cross-referencing the comitative. The verb form may be diagrammed as:
nam be
( 1 ) (2)
ba da
(3) (4)
kur . e 0
(5) (6) (7)
( 1 ) nominal element of compond verb
(2) modal-prefx
(3) conjugation-prefx
(4) comitati ve dimensional-prefx
(5) verbal element of compound verb
(6) mam-sufx
(7) personal-afx cross-referencing agent (Gilgamesh).
Lesson 12
Discussion: structure
The stucture of this text is:
[Bilgame.Endimgig. (ak), luga1.ani] .(r)
[Urammu, nitab.kalaga, luga1.Urim.a(k),
luga1.KiengLKiuri.k] .e
[ud e.Nanna. (k). 0 mu.(n.)du.0.a] .a
[nam.ti1.ani].e
a.mu.na.(n.)ru
[Iu musara.bLa u.bi .b.ur.e.0.a] . (d)
[Bilgame] .e
nam. be. ba.da.kur. e. 0
- Phonology
benefactive
agent
time
purose
verb
accompaniment
agent
verb
137
The second element of the compound verb nam ... kur
5
is variously transliterated as
kU
5
' kud, and kuru
5
(the latter with an overhanging vowel). The Idl - Irl alteration helps
show that the (amissable) Auslaut of this root was probably some kind of Irl sound which
did not exist in Akkadian. The Akkadian writng system sometimes reproduced it by Irl,
and sometimes by Id/. Moder Sumerologists sometimes tansliterate it as d, c, or t. It
seems to occur as the last consonant in some dozen or so verbal roots; another instance
occurs in Text 17. Its presence as the frst consonant or medial consonant of a root is much
harder to detect.
The kur
5
-sign can also be read tar. Confusingly enough, there appear to be two
different verbs: nam-kur
5
'
meaning "to curse", and nam-tar, "to decide the fate of If or", "to
decree a destiny for". The boundary between the two expressions is, however, sometimes
unclear, and occasionally transliterations are less than precise in differentiating between the
two; both expressions are sometimes found tansliterated as nam-tar.
- Moods
The morhology of the moods in Sumerian is quite complex. The single most
important work to unravel them is Dietz Otto Edzard 197 1ff; this is a series of articles
which it pays to keep close at hand. Some of Edzard's conclusions were modifed by
Burkhart Kienast ( 198 1 b).
-Conjugation-prefxes
In the paradigms presented in this book, model verbs in the bamt! are generally cited
with the conjugation-prefix mu, and those in the maru are generally cited with the con
jugation-prefx 1. While this does represent the most common distribution (at least in the
Ur III royal inscriptions), it is also possible to fnd verbs in the bamt! with the con
jugation-prefx 1 and verbs in the maru with the conjugation-prefx mu (although this latter
is rather rare).
It is sometimes stated that the conjugation-prefx ba "represents" bi with an additional
locative marker of some kind. It is difcult to say whether or not such a statement is
138 Lesson 12
correct. It is a fact, however, that when the conjugation-prefx ba is present, there is
frequently a locative phrase somewhere in the sentence. In purely synchronic terms, bi and
ba occupy the same slot. At least in the Ur II texts, they do not co-occur (that is, they do
not appear together in one verbal chain). It has, however, been claimed that there are cases
of ba and bi co-occurring in the somewhat earlier Gudea texts.
The conjugation-prefx bi differs from mu, i, and ba in that the only dimensional
prefx which can follow it is the locative dimensional-prefx .ni. The reasons why are
unsure.
Some scholars believe that writings such as bi -ni should be read as bi-i, with two
conjugation-prefxes. And as was said above, it has been speculated that ba and bi may co
occur in the same one verbal phrase in the Gudea texts. Such interpretations, which state
that it is possible for more than one conjugation-prefx to co-occur within one verbal
phrase, run counter to the view presented in this book, that only one conjugation-prefx can
so occur.
-Comitative -patient
In the analysis given here, lines 1 2- 1 3 are in the comitative case. However, the
apparent absence of the case-marker da does give one pause. Some people believe that
lines 1 2- 1 3 are the direct object of nam . . . kur
S
. The comitative dimensional-prefx would
represent what Poebel called the "erstarter Gebrauch des Infxes", that is, "frozen use of
the [dimensional] infx". In Gragg's study of the dimensional-prefxes in Sumerian literary
texts, he found a large number of instances where a comitative dimensional-prefx ocurred,
without any corresponding comitative case-relationship.
This is not impossible. However, there is a more general issue here. In the compound
verb nam . . . kur
S
'
nam is apparently a (historic) direct object (patient) of kur
S
. Now, it is
usually believed that Sumerian does not permit two patients in one sentence. In cases
where one might expect two patients, one of them will be expressed through one of the
adverbial cases. In the immediately preceding lines, for instance, the English tanslation
was "to erase this inscription". The direct object (patient) in Sumerian was u; the direct
object in English, "this inscription", was expressed via a loative: "to move one' s hand
over". If it is true that Sumerian does not tolerate two patients, and if it is also true that this
rule applies to historic direct objects (patients) of compound verbs, then lines 1 2- 1 3 must
be marked by another case. Given the presence of the dimensional-prefx da in the verbal
chain in line 15, this would most likely be the comitative case.
The argument that Sumerian does not permit more than one patient in a sentence is
based on general linguistic theory, and on empirical observations in Sumerian. Some
linguists would say that no language has more than one patient (at least in the deep
structure); if two seemingly occur, one must be in an adverbial relationship. However, it is
not clear if such a constaint would apply to compound verbs. Even though the frst
element of many compound verbs is historically the patient of the verb, it is not always sure
ifit functioned as such in historic Sumerian.
Lesson 12 139
-Conjugation
Following is the paradigm for the mam of the transitive verb in the singular. The
model used is sar, which is a member of the afxation class. The conjugation-pref used
here is 1.
frst person singular
second
third
1-sar-re-en
i-sar-re-en
i-sar-re
Lsar. e.en
Lsar.e.en
Lsar.e.0
The frst and second persons singular are identical in form. The fnal 1nl often dos
not show up in the writing.
If the rot ends in a vowel, there is fequent assimilation of the I el of the marO-sufx
into the vowel of the root.
Some Sumerologists analyze the morphology of these endings as: .e.n, .e.n, .e.0.
That is, the frst and second person markers are .n, not .en. This is a thorny issue, which
cannot be resolved here.
For verbs of the reduplicating class (the model is gar, "to place"), the forms are:
frst person singular
second
third
1
-
R
-g

-en
l-g

-g
-en
i-g

-g

Lgaga.en
Lgaga.en
Lgaga.0
This interpretation of the maru of verbs of the reduplicating class is essentially that of
Yoshikawa. However, all such reduplicated forms end in a vowel, which is subject to
contraction with the lel of the ending .en. This means that in texts, such forms as the
following are encountered: i-M-M--en, i-gf-gf-an, i-gf-M-, etc. Other Sumerologists
have argued that such writings indicate that Yoshikawa
'
s analysis is incorrect.
The marker .0 in such marO forms as Lsar.e.0 has been treated here in two ways.
First, it was called a personal-afx, cross-referencing the tansitive subject. Second, it was
called a marker for the third person. As discussed in Lesson 1, these are not contadictory
interpretations, but are rather two ways of saying the same thing.
-marO formation
It was Yoshikawa who established the three clases of mam formation discussed
above. Edzard has expanded this into fve classes:
( 1 ) "Unchanging": The hamtu-root and the maru-root are the same. This cor-
responds to Yoshikawa
'
s "afxation" class; Edzard does not believe that .e is a marO
marker.
(2) "Reduplicating": Same as Yoshikawa.
(3) "Root-varying": The two roots are different, but similar phonetically: "To
approach" is te in the bamn!, but teg in the mam and teg are the same sign).
(4) "Replacement": Same as Yoshikawa.
(5) "Iregular": These do not seem to ft nicely into the other four categories.
140
Lesson 12
It is not yet clear whether Y oshikawa
'
s or Edzard's classifcation scheme is to be
preferred. As progress in Sumerology is made, it is probable that more classes, and fner
subdivisions within these classes, will have to be made. Yoshikawa himself has indicated
that his scheme needs to be expanded; Kienast ( 198 1 b) has suggested some modifcations
to the scheme of Y oshikawa and Edzard. In this book, Y oshikawa's system of
classifcation into affxation, reduplication, and alteration classes has been followed,
because this classifcation scheme works well for the Ur III royal inscriptions.
- Roots
As may have been inferred above, it is only in the last ten or ffteen years that the
morphology of the maru has become somewhat clear, thanks primarily to the work of
Yoshikawa and Edzard. Even now, however, there remain thory problems. Many
difculties are occasioned by the fact that in some cases the same sign stands for two
different roots, one for the bamt!-root, and one for the mara-root (such as girdu men
tioned above). The situation is still more complicated, however; certain roots appear to
have separate forms for singular subjects and for plural subjects, in both the bam! and the
mara; this produces at least four different roots. There appear to be cases where certain
roots have a bam!-singular root, a mara-singular root, and a different plural root used for
both the bamt! and the mara. In at least one case there appears to be a root which has
different forms, depending on whether the object is singular or plural.
This complexity is to be expected; other languages of the world show such diversity in
morphology. As progress is made, more such cases will be identifed.
- Moods
There has been much discussion about the precise etymology of the Akkadian
grammatical terms bam! and maru, and even more discussion about the distinctions which
are marked by these two terms. The difference in function between the bam! and the mam
has been variously seen as a difference in tense, or a difference in aspect, or a difference i
Aktionsart. In the Ur III royal inscriptions, they seem more tense-like than aspect-like; the
bamt! is regularly used for past action, and the man} for future action. However, when
dealing with more complicated texts, especially literary texts, such a single binary
distinction will not work. The fact that certain modal-prefixes require either the bam! or
mam, seeming I y regardless of tense, complicates the issue.
-Ergativity
The term split ergative has been applied to Sumerian, because the personal-afxes
behave in an ergative way in the bam!, but not in the maru; in the man}, they behave in a
nominative-accusative way. Consider the following sentences:
( 1 ) The king built the house .
A
lugal.e e.f mu.n.du.q
(2) The king went.
lugal.f mu.gin.
Lesson 12
(3) The king will build the house.
+
lugal.e e..du.e.0
(4) The king will go.
lugal. Ldu.(
141
In (3), the direct object is cross-referenced by the personal-affx in the pre-verbal rot
slot. In (4), the subject of the intansitive verb is cross-referenced in the post-verbal rot
slot. Since the direct object in (3) and the intansitive subject in (4) are not cross-referenced
in the same manner, they cannot be considered to function in an ergative way, but rather in
a nominative-accusative way.
The case-markers in (3) and (4) - and in ( 1 ) and (2) -are the same; the difference is i
the way that the case-markers are cross-referenced in the verbal phrase. In ergative
languages which lack a case-system, ergativity only shows up in the cross-referencing
system. Discussing a Mayan language called Sacapultec, for instance, Du Bois says: "As
in all Mayan languages, the ergative pattering of Sacapultec morphology is entirely in the
verbal cross-referencing infection; nouns are not case-marked for grammatical relations"
( 1 985: 809).
Split ergative languages help demonstate that the ergative - accusative distinction is
more of a continuum than as a simple dichotomy. Du Bois has suggested that split
ergativity is motivated by pragmatic or discourse pressures, forcing a re-alignment of the
principal constituents of a sentence.
The explicit statement that Sumerian is a split ergative language has been particularly
advanced by Michalowski. Not everyone is in agreement with these views; some scholars
believe that there are too many apparent exceptions which cannot yet be explained.
-Reduplication
Reduplicaton plays many roles in Sumerian. In this Lesson, reduplication is one of
the devices used to derive maru-roots from bamm-roots. A second common use is often
referred to as "free" reduplication, or "bamm" reduplication. This consists of reduplication
of the bamm-root; for example, l
-
gm-ga. This is not a case of mara reduplicaton, because
the mara of gar is
i-M-
g
.
The functions of free reduplication are not all clear. In many ways, it appears to be the
functional equivalent of the D-stem ("Intensive") in Semitic. Just as it is not easy to
categorize all the uses of the D-stem in Semitic, so it is not possible to easily categorize all
the uses of free reduplication. Edzard
'
s preliminary classifcation includes such things as:
stressing of plural or totality of subject or object; plurality of occurrences; distributive
relations; etc. In bilingual literary texts, reduplication is ofen translated by the Akkadian
Gtn ("Iterative") ster.
Free reduplication is not uncommon in Sumerian; for example, it occurs frequently in
Gudea and in Old Babylonian literary texts. There appear to be no cases among the Ur III
royal inscriptions, perhaps simply because of content; plural objects do not appear to be
mentioned.
Certain verbs seem to have become lexicalized in the form of a reduplicated rot. For
1
4
2
Lesson 12
example, .l 4 -ur4
'
"to look for and gather up", almost always appears this way.
Presumably, this is because it is almost always used with a plurality of objects. Such a
lexicalizaton has a parallel in the Semitic languages, where certain roots are lexicalized in
certain stems. For example, the Akkadian buu, "to look for", only occurs in the D-stem.
I context, it is often very diffcult with certain verbs to decide whether maire
duplication or free reduplication is present. And if it is a case of fee reduplicaton, it is
often not easy to see its function.
It is probable that reduplicated forms (of whatever kind) were phonetically reduced in
speech. There is evidence for this fom texts written in syllabic orthography (Appendix 2)
and from unusual occurences of syllabic writings within normal orthography. For
example, an expected biz-biz- appears once in Gudea as: bi-bi-ze. Few details of such
phonetic reduction are understood.
-Origin of cases
It has been speculated that the Sumerian case-endings (some of them, at any rate) were
originally nouns. In particular, it has been claimed that the comitative case-marker da is in
origin the same da meaning "side", seen in the expression "king of the four quarters",
lugal-an-ub-da-limmuba. This is not impossible, but it is harder to fnd an etymology for
the other case-markings.
-Research in Sumerian
It is sometimes fairly easy to understand the meaning of a Sumerian text, and even
relatively easy to describe, on surface terms, what we see, but it is much more difcult to
understand exactly what is happening. For example, based on context and on parallels in
Akkadian, Phoenician, and Aramaic texts, all Sumerologists would understand the last lines
of this text to mean, "May Gilgamesh curse the man who erases this inscription",
regardless of the presence or absence of a comitative case-marker, regardless of the
presence or absence of a dimensional-prefx, regardless of the distribution of the personal
affxes, etc. However, unless such details are well understood, it is much more difcult to
fgure out the meaning of really complicated passages.
This text also illustrates the problems encountered in doing research in Sumerian. An
obvious question which arose when discussing the presence or absence of a comitatve .da
at the end of line 13 was, "How is this verb construed in its other occurences? Does it use
the comitative or some other constuction?" Unfortunately, without an up-to-date Sumerian
dictionary, such questions are not easy to answer. One can look at the existing dictionaries,
or texts with glossaries, but without a painstaking examination of many sources, it is
impossible to be sure that all instances of any particular word have been found.
-Curse-formulae
It is not uncommon for votive inscriptions to be provided with a curse-formula.
Typically, the frst part of an inscription will form a straight-forward text; the curse is
tacked-on at the end. In the Ur III texts, only a limited number of curse-formulas ocur; the
next occurence is in Text 15.
Lesson 12 143
-History
The historical Gilgamesh was the ffth king of the First Dynasty of Uruk, which falls
within the Early Dynastc IT period (about 2700-2500 BC). No inscriptions of his are
preserved, or contemporary references to him, but there are a few inscriptions of his
approximate contemporaries.
The frst attestation of him is in a god-list frm Fara, where his name is written:
d
BiI
PAP-g-mM. It is difcult to say exactly how these signs represent the name /BilgameS/;
his name is spelled several different ways in the course of Mesopotamian tradition.
Gilgamesh seems to have been very popular with the Ur III kings; according to Jeffrey
Tigay, "The kings of Ur ITI regarded Gilgamesh as something like their personal god"
( 1 982: 1 3 n.50). Michalowski speculates that the "Gilgamesh stories were made par of the
school curriculum during the Ur III period" ( 1 987: 66). In his hymns, Shulgi refers to
Ninsun, the mother of Gilgamesh, as his own mother, and he refers to Gilgamesh as his
brother.
144 Lesson 12
Text 1 2a
supplemen tay
Another door socket.
5
I O

'
'

"
'

+ '

{
-

'

_ +
`

| |

'

.
|

'

'

~
Lm8on!2 I4
Note.

-dur-an-k The name of the ziggurat at Nippur. Etymologically, "the temple, the bnd
of (=between) heaven and eanh", E dur.[an.kiJ.k.
Text 12a was found outside of a contolled archaeological COnlext. The following
photograph is of a dor socket baring another copy of the same inscription, found in situ:
THE INSCRIPTION THAT IDENTIFIED THE TEMPLE Of INANNA
A door socket. this granite block bore a door post in its cavity. With the discovery of the thick buttressed
wall we had only the outside of a monumental building. To get inside. two plckmen dug a five.yard square
shaft beside the inner wall face. Walls and floors were soon located, but the room was larger than the shaft.
TIle few objects found did not identify the building. To lear more about the area with a minimum of
effort, we tunneted atong one wall for twenty feet, then another (or thirty feet to a corner where there was
a doorway. We dug through this, first finding a disappointingly uninscribed door socket, then. in a brick
box tower down, another with cuneiform wedges. The writing is that of king Shulgi, second ruler of the
Third Dynasty of Ur. commemorating his rebuilding of the temple of Inanna. With this, we had the cult
spot of Inanna at Nippur.
Lesson 1 3
This text is a standard inscription of Amar-Sin, the son and successor of Shulgi; he
ruled from 2046 to 2038 BC. The text exists in many copies; Text 1 3b is a stamped brick.
Sign-list and vocabulary

Zuen Zuen (DN, masc)


o * _ Amar-
d
Zuen Amar-Sin (PN)
Nibru Nippur (ON)
amar young bull
sag-us supporter, sustainer, patron

pad () to fnd, call, reveal



mu . . . pad (w to propose

I
Notes
Zuen This seems to be another name of Nanna, although it is not clear why he had two
names. Jacobsen thinks that the term Nanna refers specifcally to the god
'
s role as the "full
moon", and Zuen refers to his role as the "crescent moon". It has also been suggested that
Zuen is the Akkadian equivalent of Sumerian Nanna; that is, they are two different names
for the same deity. However, there is no obvious Semitic etymology for Zuen.
The Akkadian equivalent of this DN is usually tanscribed as either "Sin" or "Suen".
The Sumerian word was also borrowed into Akkadian as a common noun, appearing as
sInu, suenu, sinnu, and innu. It is glossed by the CAD as: "1 ) the moon 2) crescent
shaped or semi-circular object".
The writing is discussed below.
Amar-
d
Zuen Etymologically, "young bull of Zuen", amar. Zuen.(ak). The name of this
ruler is often transcribed "Amar-Sin" or "Amar-Suen", which are really Akkadianized
transcriptions.
Early scholars believed that this PN was Akkadian. The Akkadian equivalent of amar
is bOru. Therefore, the name appears in some older secondary literature as: "BOr-Sin", or
something similar. Almost all moder scholars believe that the name is Sumerian, although
there is really not much evidence to prove this.
147
148 Lesson 13
Nibru One of the more ancient and important cities in Mesopotamia; Kramer has called it
the "spiritual and intellectual center of Mesopotamia". J acobsen has said:
From the very beginning of historical times Nippur and Enlil were recog
nized as an undisputed source of rule over Sumer as a whole, and kings of
Sumer would derive their authority from recognition in Nippur rather than
fom their own city and its city-god ( 1957: 1 39).
The modern name of the site is Nuffar. It was the frst tell to be excavated by
American archaeologists (the University of Pennsylvania, beginning in 1 887). Over thirty
thousand tablets were found, mostly in Sumerian, ranging from the third to the frst
millennium BC. Large numbers of these texts are still unpublished. The vast majority of
our Sumerian literary texts are Old Babylonian copies found at Nippur.
The etymology and writing of the name are discussed below.
sa
g
-us The exact etymology and meaning are not sure. , as seen previously, means
"head". us has several meanings. It often means "to lie against", "to lean against". (It is
used as a logogram to represent Akkadian emedu, "to lean against, to reach".) It may also
have a transitive sense, "to lift", i.e., "to support". Here it may be an active participle in .0,
with sag being its historic patient or incorporated object: "the one who lifts the head". The
usual translation is "supporter, sustainer, champion (of)". It is translated by Sollberger as:
"protector, patron (literally, '(he who) supports the head')".
Lesson 13
Text 1 3a
?

|
M
.
m
= =

|
[
|
8 8 W
r-i l
t f l *
1
44
o
<

H
=
i [
|
l
'|

| |
_

t
H
=

@ @

=. W M
. . .. - - -I
w u m
M >

-I

,If'


'
F
149
There is a certain amount of variation in the cuneiform signs in this text. The da at the
end of line 4 has an initial vertical, not seen in the da in line 9. Similarly, the sign read nitab
in line 7 has a vertical not seen in the sign read us in line 5.
150
Lesson 13
Text 1 3b
1 :
2:
3:
4:
5:
6:
7:
8:
Transliteration
d
Amar-
d
Zuen
Nibru
ki
_
d
En-lil-Ie
mu-pad-da
sag-us
_
d
En-lil-ka
nitab-kalag-g
I I U '
ki
.!!- nm
S
-ma
Lesson 13 1 51
Transcrption Translation
Amarzuen Amar-Sin,
Nibru.a
|
roosed
Enlil.e
y nlil
in Nippur,
mu.pad.a
sagus. patron of
e.Enlil.(a)k.a(k) the temple ofEnlil,
nitabkalaga the mighty man,
lugal.Urim.a(k) the king of Ur,
9: lugr-an-ub-da-liruzbalugal.anub.da.(k) limmu.bLa(k) the king of the four quarters.
Comenta
1. As was the case with Shulgi, the name Amar-Sin is preceded by the determinative for
divine names. There are thus two divine deteninatives in the line: the second is for the
DN Zuen, and the frst is for the PN Amar-
d
Zuen.
2-4. These lines form a relative clause modifying Amar-Sin of line 1 . mu means "name",
and pad is something like "to reveal". mu . . . pad is a compound verb, meaning approx
imately "to propose". The underlying idea is that the name of Amar-Sin was proposed by
Enlil in the council of the gods, meeting in Nippur, to be the king of Sumer and Akkad.
Sumerian has two ways of forming relative clauses. The frst has occurred several
times: it consists of nominalizing a complete independent sentence by .a, and then placing
this nominalized sentence in apposition to a relative marker. This is sometimes called the
"full" relative clause. For example, "He built the temple of Nanna" is: e.Nanna.(k).0
mu.n.du.0. "The one who built the temple of Nanna" is: lu. [e.Nanna.(k). 0 mu.n.du. 0] . a.
In this example, the relative marker is logically the subject of the verb in the relative
clause: "the man who built", "the man who shall erase". However, oblique relations are
also possible, such as: "the temple which the king built", "the god for whom the king
dedicated a votive offering".
In a clause such as "the temple which the king built", "temple" is logically the direct
object of the verb. In such relative clauses in Sumerian, there is no overt marker indicating
this relationship. This clause could be expressed as: lugal-Ie mu-du-. The simplest way
to understand this construction is to think of it as: "[the temple] [the king built (it)]". In
English, there is no overt marker, except in the distinction between "who" and "whom":
"the man who built", but "the man whom he saw". In the classical Semitic languages, a
resumptive pronoun is used. For example, the Akkadian equivalent of "the temple which
the king built" is: btu a arru ibmlu. Literally, this is: "the temple which the king built
it". Sumerian uses no relative marker of any kind (unlike the use of English "that, which",
or Akkadian a), and there is no overt marker for the direct object (unlike the Akkadian u).
The second way relative clauses are foned in Sumerian is less understood in all its
details. It is sometimes called a "reduced" relative clause, and sometimes a "participial
152
Lesson 13
constuction". It is foned by deletion of the entire verbal prefx chain, and nominalization
of the remaining verbal root in .a (except in certain cases, not discussed here).
To express "Amar-Sin, whom Enlil proposed in Nippur" using the full form of the
relative clause, would be approximately: Amarzuen [Enlil.e Nibru.a mu.mu.n.pad.0]. a.
(The frst Imul is the nominal component of the compound verb mu . . . pad: the second
Imul is the conjugation-prefx.) To express the same idea using the reduced relative
clause, the prefx chain is deleted: Amarsin [Enli1.e Nibru. a mu.pad.0].a, which is the fon
in Text 13. The mu which remains is the nominal element of the compound verb, not the
conjugation-prefx. Since the nominal component of a compound verb is not part of the
prefx chain, it is not deleted.
Several English tanslations of this constuction are possible: "Amar-Sin, whom Enlil
proposed in Nippur", "Amar-Sin, proposed by Enlil in Nippur", etc.
Both full and reduced relative clauses are common in Sumerian, but it is not known if
there are rules govering their distribution. Certain fonulaic expressions tend to prefer
one constuction, while other fonulaic constuctions prefer the other. For example, "the
man who built . . . " always appears in these texts as lu in-du-, but "the man proposed by
Enlil" always appears in these texts as lu
d
En-lil-Ie mu-pad-da.
5. sag-us _
d
En-lil-ka sagus.e.Enlil.(a)k.a(k), "patron of the temple of Enlil". The geni
tive phrases seen till now have consisted of two nouns or nominal phrases. However, it is
also possible to have a genitive phrase consisting of three or four elements, such as "the
king of the temple of N anna". Such genitive phrases are foned by the addition of an extra
". ak" for each new element in the genitive phrase. For example, "the temple of N anna" is:
e.Nanna.k; "the king of the temple of Nanna" is: luga1.e.Nanna.k.ak; "patron of the temple
of Enlil" is: sagus.e.Enli1.ak.ak. (Sequences of two ".ak"s are sometimes referred to as
"double genitives".)
Sequences of four nouns or nominal phrases (therefore, with three ". aks"), although
penissable, are uncommon. Sumerian does not seem to tolerate a sequence of more than
three ". ak"s; if such a situation would arise, no more than three are used. More commonly,
a periphrasis of some kind is used instead.
The genitive phrase in this line is written sag-us-_
d
En-lil-ka. The expected la! of the
frst genitive marker does not appear in the writing. One might have expected a writing
such as sag-us-_
d
En-lil-Ia-ka, or some such. A similar writing occurred in Text 1 1 : bi-
l

nam-munus-ka-ni, for bili.nam.munus. (a)k.ani. As discussed at length in that Lesson, it is


not known whether the problem is at the orthographic, phonological, or morphological
level. Falkenstein
'
s school, for example, would read the lil-sign here as lila
2
: sag-us-
d
En-lil! ka.
Discussion: structure
Text 1 3 is a standard inscription, similar to Text 7. It consists entirely of a series of
appositives, serving as epithets to the name Amar-Sin in line 1 ; there is no fnite verb form:
Amar-Sin,
proposed by Enlil in Nippur,
patron of the temple of Enlil,
the mighty man,
the king of Ur,
the king of the four quarters.
-Orthography
Lesson 13 153
The name Zuen is composed of two signs, the en-sign and the zu-sign. The two signs
are always written in this order. However, there is a fair amount of evidence which shows
that the zu-sign was actually pronounced before the en-sign; that is, this name was
pronounced something like /zuen/. For example, the Akkadian word sInu (discussed
above) is a loan-word from Zuen. In the bilingual texts from Ebla, the Sumerian version of
this divine name appears once as En-zu, but twice as En-zi. In all three cases, the Eblaite
equivalent is Zu-i-nu (read by some as Su-i-nu). Also, there are Akkadian personal names
formed with the Akkadian version of the name, which appear in Hebrew, Greek, Latin and
later English transcriptions; thus, the name Sin-ahhe-erba ("Sin has replaced my [dead]
brothers") appears ultimately i English as "Sennacherib". There are, however, some
unusual spellings in later Sumerian texts which may indicate that at times the name Zuen
was "read as written", that is, read as /enzu/ and not as /zuen/. These late writings may
result from misunderstandings of scribes.
It is not known why this order of signs is used; a similar phenomenon is discussed in
Lesson 14. In very early Sumerian, it was possible for signs within a line or case to be
written in a rather fee order, not always corresponding to the order of signs as they were
read or pronounced. It is possible that writings such as En-zu represent survivals fom this
period.
It is diffcult to decide how to tansliterate such spellings. Some Sumerologists
transliterate sign-by-sign; thus, En-zu in this case. This practice is not common in modern
day transliterations, and is regarded as somewhat old-fashioned. Other Sumerologists
follow a convention whereby the transliteration gives the signs in the order-as-read,
separated by a colon: Zu:en. Some Sumerologists will then follow this with the trans
literation of the signs in their order-as-written, usually in caps, within parentheses:
Zu:en(-ZU). The colon is sometimes used in this way when transliterating very early
Sumerian, when the order of signs in the line or case does not correspond to the presumed
order of pronunciation.
-Writing system
The Sumerian pronunciation of the name of the city of Nippur is known from lexical
lists, where En-m
ki
is spelled out as Ni-ib-ru. Similarly, the Akkadian pronunciation of the
city name is also known from lexical lists, where it is spelled out as Ni-,l-p! -ru.
The Sumerian writing of the place name represents a not uncommon instance where
the writing system tells us nothing about the pronunciation of the place-name. The
etymology of Nibru is unknown; it is presumably a pre-Sumerian substrate word.
However, the city eventually became especially associated with the god Enlil. Therefore,
the name of the city was written with the same two signs as in the god's name, but followed
1 54 Lesson 13
by the deteninative for place: En-lil
ki
. That is, the writing does not attempt at all to repro
duce the phonetic sequence INibru/. Rather, the Sumerian reader would understand the
written signs as standing for "the place assoiated with the go Enlil", that is, Nibru.
Transliterations of such place names (and of similar common nouns) vary. Older
practice tends to reproduce the basic value of the signs foning the word: Ep_li l
ki
. More
current practice is to use the name of the GN (assuming it is known) : Nibru
k
I.
-Relative clauses
The ten "participial constuction" has been used to describe constuctions such as in
lines 2-4, because a reduced verbal fon with a nominalizer is fonally identical with what
has been called here a passive pariciple; a reduced mu-n-du, becoming du-, is fonnally
identical with the passive pariciple du-. I origin, in fact, passive participles are all
probably reduced relative clauses, in special syntactic environments.
The constuction called here "reduced relative clause" is often referred to as the
"Mesanepada (or Mesannepadda) constuction". (This name for the construction
apparently goes back to Falkenstein. ) Me-an-ne-pad-da was the founder of the First Dy
nasty of Ur ("Ur I") sometime around 2550 BC. His name means "the young man whom
An chose", or "the young man chosen by An": Me.an.e pad.a. This is the minimal fon
of the construction: a head noun (mM); an agentive marked in .e (An-; a verbal root
(pad); a nominalizer. (A few other names of the type "X-an-ne-pad-da are also known.)
Lines 1-4 of Text 13 are only slightly more complicated; Text 13 includes a locative phrase,
and also uses a compound verb (mu . . . pad) instead of just pad.
Other examples of this constuction ocur in Gudea. A temple is referred to as: E
ninnu An-ne ki-gar-ra, "the Eninnu temple, which An established (literally, 'placed on the
ground
'
, ki.a)", or "the Eninnu temple established by An".
The only study that deals specifcally with relative clauses in Sumerian is by Gragg
( 1 972a). It was written for a non-Sumerological audience, and is by design short and
schematic, but has several useful observations. He points out that the syntax of relative
clauses, particularly reduced relative clauses, i s not completely understood. Henri Limet
( 1 975a) has studied the parallel use of .a in participial and relative sentences.
-Standard inscriptions
Copies of this text have been found at several different sites, including Adab, Bad
Tibira, Eridu, Girsu, Isin, Kisurra, Sippar, Tell el-Lahm, Ur, and Uruk. All copies were
inscribed on bricks. This situation is not uncommon with standard inscriptions, which can
be found anywhere the ruler held sway, or where building-activity was conducted under his
aegIS.
- History
Shulgi apparently died of old age; he had ruled for some 47 years. His son, Amar
Sin, ruled only nine years, and not much is known of his activities.
Lesson 13
155
Text 1 3c
supplementa
Another brck.
Lesson 1 4
This text is a brick building inscription of Amar-Sin; it is essentially an expansion of
Text 1 3.
Sign-list ad vocabular
En-ki Enki (DN, masc)
Notes
abzu "apsu", water basin
En-ki The god of the subterranean waters, and also the god of wisdom. He was a son of
Nammu. His name apparently means "lord of the earth", en.ki.(k). There are spellings
which show that this name is a genitive phrase, not a noun-noun compound.
Why a god whose name means "lord of the earth" became assoiated with water is not
entirely clear; Jacobsen has spoken of "the role of water in fuctifying the earth". It has
also been speculated that the element Ikil appearing in this name is a different word than
the word lkil meaning "earth", perhaps the same ki appearing in the compound verb
ki g
2

The Sumerian god Enki was equated with the Akkadian god Ea <-. The name of
the latter is of uncertain etymology; it does not infect for case. In the bilingual lists from
Ebla, the Ebaite equivalent of Enki is written E-!
9
' This would appear to be an infected
form of the name, with the nominative case-marker. It has also been speculated that the
Akkadian writing E- and the Eblaite writing E-!
9
are actually phonetic spellings repre
senting a Semitic form something like Ibayyu/, "the living one". This idea is explicitly
developed by Cyrus Gordon ( 1 987: 1 9-20).
abzu This is composed of two signs, the zu-sign followed by the ab-sign. However, it is
known that the ab-sign was read before the zu-sign. That is, the word was pronounced
something like labzu/. The phenomenon is similar to that of the DN Zuen, which was
pronounced lzuen/, although written en-zu. In older tansliterations it may appear as zu
ab.
The original use of this term was mythological. It referred to the subterranean fresh
waters, which the Sumerians believed lay below the surface of the earth. These waters fed
the wells, streams, rivers, marshes, etc. These waters were the special purview of Enki.
The term abzu was also used as the name of a large temple in Eridu, built to honor
Enki. Most of the work on this temple was done by Amar-Sin, although it was his father
who actually began the construction. This temple apparently stood over a fresh-water
lagoon.
The term was later applied to a cultic object, presumably some kind of water-basin
1 57
158 Lesson 14
used in the temple. At a number of sites in Mesopotamia, objects have been found which
archaeologists have identifed with the term abzu. Several such objects have recently been
found at Ebla.
This word was borrowed into Akkadian as apsf, glossed by the CAD as: "1 ) deep
water, sea, cosmic subterranean water, 2) (a personifed mythological fgure), 3) water
basin in the temple".
It is possible that the Eblaite equivalent of Sumerian abzu and Akkadian apsf appears
as -ba-si (Fales 1984: 1 84), but the interpretation is somewhat uncertain.
The English word "abyss" is thought to derive from this Sumerian word, via
Akkadian and Greek. It has been speculated that the word is not native Sumerian, but
rather derives from a substrate language.
Lesson 14
Text 1 4
-'l
.>

+I

^ - "
-
'
----
<'
159
160
1 :
2:
3:
4:
5:
6:
7:
8:
9:
Transliteration
d
Amar-
d
Zuen
d
En-lil-Ie
Nibru
ki
_
mu-pad-da
sag-us
_
d
En-li1-ka
lugal-kalag-gi
I I U
' ki
..gl
-
nm
S
-ma
lugil-an-ub-da-limmu
I
ba-ke
4
1 0:
d
En-ki
1 1 : lugal-ki-g-M-ni-i
1 2: abzu ki-g-M-ni
1 3: mu-na-du
Comenta
Lesson 14
Transcripton Translation
Amarzuen Amarsin,
Enlil.e _roosed
Nibru.a
y nlil
in Nippur,
mu.pad.a
sagus patron of
e.Enlil.(a)k.a(k) the temple ofEnlil,
lugal.kalaga the mighty king,
lugal.Urim. a(k) the king of Ur,
lugal.anub.da. (k)
limmu.bi.ak.e
the king of the four quarters -
Enki for Enki,
lugal.ki.aga.a.ani.r his beloved king -
abzuki. aga.a.ani.0 his beloved apsi -
mu.na.(n.)du.0 he built.
9. lugal-an-ub-da-limmu
T
ba-ke
4
'
Cf. line 9 of Text 1 3: l!gal-an-ub-da-limmu2 -ba. The
difference between the two is the presence of the ergative case-marker in Text 14. It was
not present in Text 13, because there was no fnite verb form in that text; rather, Text 1 3
consisted of a string of appositives. But i n Text 14, all the appositives are part of the
nominal phrase expressing the agent of the transitive verb in line 13. As stated in Lesson 1 ,
the nominal phrase to which the case-markers are attached in Sumerian can vary
considerably in size - all the way from a single noun, to long complexes such as this one: a
nine-line nominal phrase.
1 1 . The dative case-marker Irl is here expressed. Its occurrence in the royal inscriptions
of Ur-Nammu is unsure, and it does not appear in the inscriptions of Shulgi. At some
point in the reign of Amar-Sin, there was apparently a change in orthography, although the
motivation for this full writing is unknown. During the time of Amar-Sin there are wri
tings with the dative case-marker Irl expressed in the writing, such as here, but there are
also texts where it is not expressed.
The problem cannot be described simply in chronological terms. Even in the Gudea
texts, there are isolated instances of the Irl appearing in the script. For example, "to the
king" is normally written either l!gal-i-ni or l! gal-ni in Gudea, but l!gal-ni-ir occurs at
least once. "For Gudea" is written: Gu-de-; the spelling Gu-de-i-ar is also attested. Fal
kenstein, in his study of the Gudea inscriptions, could fnd no rules governing either the
morphology or the orthography of the dative.
It has been argued that the problem here is phonological, not orthographic. Poebel and
Falkenstein have suggested that in a writing such as l!gal--ni for the dative, the original
Lesson 14 161
Irl of the dative case-marker was completely lost, proucing /lugalnil for the dative.
However, adoption of such a view entails rather baroue convolutions i explaining such
forms as hgal--ni-ir. Falkenstein, in fact, speaks of "eine sekund1e Restitution" of the
dative case-marker Irl, refecting a perio when Sumerian was beginning to fall out of use
as a spoken language. This would mean that the original form was Irl, but then the Irl
was lost, and later it was "restored". However, such an explanation encounters strong
linguistic objections, and also historical objections. Also, it does not account for the rise of
explicitness seen in other areas of the grammar. In Text 1 6, for example, there occurs mu
na-an-du, the frst occurence in the boy of texts studied here where the personal-affx
crss-referencing the bamt!-agent is actually written.
12. Because of the ambiguity of the term abzu, it is not clear what the patient in this line
refers to: the temple built by Amar-Sin and his father, or to a cultic object within this
temple.
Discussion: structure
The structure of this text is:
[Amarzuen, Enlil.e Nibru.a mu. pad.a,
sagus.e.Enlil. (a)k.a(k),
lugal.kalaga, lugal. Urim.a(k),
lugal.anub.da.(k) limmu.bLak].e
[Enki, lugal.kLaga.a.ani].r
[abzu ki.aga.a.ani].0
mu.na.(n.)du.0
agent
benefactive
patient
verb
The frst nine lines are essentially the same standard inscription seen in Text 1 3. The
rest is a straight-forward building inscription. The result is that this inscription begins with
the name of the king (the agent) instead of with the name of the deity (the benefactive). It is
as if the scribe began with a stock standard inscription, then tacked on a building
inscription.
There are two differences between Text 1 3 and Text 14. The frst is in the order of
lines 2 and 3. In Text 1 3, the locative phrase within the reduced relative clause precedes the
agentive phrase, but in Text 14, the agentive phrase precedes the locative phrase. Text 14
uses the more usual syntax. In Text 1 3, there is presumably some emphasis on the word
"Nippur". The second difference is in line 7. Text 1 3 uses a title which occurred in several
other texts: nitab kalag-g. Text 14 uses lugilkalag-g,.
-Orthography
On the one hand, writings which explicitly represent the dative case-marker Irl may be
viewed as part of the general process of the Sumerian writing system becoming more and
more explicit in its representation of phonological and morphological features. On the other
hand, it is hard to understand exactly how such a practice appeared - what motivated a
scribe, practicing by its nature a conserative craft, to write the Ir!?
It is not known when Sumerian began to die out as a spoken language; many scholars
1 62 Lesson 14
believe that such a prcess was already on-going during the Ur III period. If so, the
increase in explicitness in the texts - such as the writing of the dative case-marker Irl -
may be corelated with an increased need of the scribes for help in reading and writing
Sumerian. That is, as the scribes' knowledge grew more and more "shaky", there was a
need to write the morhemes down in an unambiguous way. At the same time, there may
have been felt a scholastic tendency to write all morhemes down. Similarly, in our own
scholastic tadition of tanscribing Sumerian, we are prone to write down full underlying
forms of morphemes.
-Ergativity
There are two ergatve case-makers in this sentence: the .e in line 9, marking the
agent of the main verb in line 1 3, and the .e in line 2, markng the agent of the verb in line 4
(embedded in a relative clause). Potentially, this could cause a certain amount of confusion.
If one thought that the .e in line 2 marked the agent of the verb coming up in line 1 3, the
text would start to become rather confused. In practice, however, the formulaic nature of
these texts helps to prevent such confusion. In the spoken language, there were probably
features such as stress and intonation which helped obviate such problems.
-Textual problems
At the end of line 9, the autograph reads ke4 ; the ke4 -sign combines the Ikl of the
genitive marker with the lel of the ergative case-marker. This particula use of the ke4 -sign
has ocurred in several of the previous texts.
However, at least two of the three other published exemplars of this text read a ka
sign, not a ke4 -sign, at the end of line 9. This is hard to explain. There is no evidence for a
(phonological) change of le/ ) lal at this period; sporadic cases do occur in later Sumerian,
but under different conditions.
One possibility is to see a long anticipatory genitive, of a kind not seen previously.
The essence of the sentence would be:
"Of Amar-Sin ... , to Enki his beloved lord" )
"To Enki, the beloved lord of Amar-Sin".
Although somewhat similar anticipatory genitives do ocur in Sumerian, there seem to be
no exact parallels to this construction. Also, such an interretation would not leave any
overt agent for the fnite verb.
It is diffcult to fnd a satisfactory explanation for the writing with the ka-sign. It is not
simply a scribal error, since the ke4 -sign seems to ocur in only one exemplar; two
exemplars clearly have the ka-sign, and one exemplar is slightly damaged at the crucial
point.
The fact that the ke4 -sign is apparent in only one copy is in itself suspicious. It is
possible that this writing represents an (unconscious) attempt by the scribe to bring the text
into line with more common Sumerian morphology and orthography. (This assumes, in
fact, that the ke4-sign is indeed present, and that it is not an error on the part of the moder
day editor of the text; there is no photograph available of the text.)
Lesson 14 163
-Titulature
Texts 1 3c and 14 use the title lugal-kalag-g, in place of nitag-kalag-g.!. According to
Hallo, Amar-Sin was the frst king of the Ur lI Dynasty to adopt this title. He and his
successors used it "to the virtual exclusion of the older title".
-Terminology
Occasionally, Sumerologists and Assyriologists will informally use an Akkadian
word, even when referring to its Sumerian counterpart. For example, even when
discussing the abzu, in a Sumerian context, Sumerlogists will not infrequently refer to "the
apsu". There is no theoretical or ideological reason for such practice; it refects the fact that
Sumerologists lea Akkadian before they lea Sumerian, and also the fact that more is
usually known about the Akkadian word and its referent than about the Sumerian word.
16 Lesson 14
Text 1 4a
supplementary
Tis was inscribed on a cone.
Lesson 1 5
This is another standard inscription of Amar-Sin. It may have been inscribed on the
pedestal of a statue.
Sign-list ad vocabula
alam statue
mu name

ama mother

numun seed; offspring, progeny
.
sig9 () to be narrow

kur to change

bU
6
to tear out, to uproot
f-
til to put an end to
1
me

Im

b


eb
Notes
alam Transliterated by some as alam, and by others as alan. Some Sumerologists believe
that alan is the older, and alam the later form. On the other hand, Lieberman reconstucts the
original form as /alag/; he believes that nasals in word-fnal position were neutralized with
respect to point of articulation. Others have postulated some kind of connecton with the
word appearing in Akkadian as almu, in Hebrew as elem, "image".
baag-sig
9
-ga The exact meaning is unsure. It may mean "narrow dais", barag.sig.a. It
is translated by the PSD as "socle (of a statue)". It was borrowed into Akkaian as
barasigQ, translated by the CAD as: "low socle for cultic purposes".
1 65
1
66
Lesson 15
kur This was equated with Akkadian aml and nakaru. Both have many meanings. For
nakaru in the D-stem, the CAD lists, among others, " . . . 8. to discard an object (tablet, stela,
etc. ), to remove an inscription . . . 9. to clear away rubble, etc. , to discard, remove from a
container, to demolish a building . . . 1 1 . . . . to place an object in a new location . . . ".
Most frequently, kur govers a direct object (patient).
bU6 This is translated by the PSD as "to tear out", "to pull out", "to uproot", "to extirpate".
Although the meaning is clear, the precise reading is not. The sign is a ka-sign with an
inscribed kar-sign. This is most clearly seen in the Neo-Assyrian form of the sign: .
Sumerologists often transliterate such inscribed signs with an "x": KxKR.
Older works transliterate the ka-sign as bus (when necessary), KxKR as and
KxSU as bu. Recent works, however, transliterate bt KxKR and KxSU as bu,
and add the fuller transliteration in parentheses: bu(KxKR); bu(KxSU). However,
such a system can lead to confusion, because it is easy for the forms in parentheses to be
accidentally dropped in the mechanical process of printing. And, since the entire function
of transliteration is to provide a one-to-one correspondence of a specifc cuneiform sign
with a specifc transliteration, it seems counter-purposeful to use bu for two different signs.
Therefore, the older procedure is followed here, and the sign is tansliterated as .
It is probable that the root of this word ended in some kind of Ir/-Auslaut; in Text 1 5
i t appears i n a verb form written i-bu
6
-re-. However, no Iburl value for this sign is re
cognized by the standard sign lists. A possible reading buzurs is recorded, but this may be
some other use. The problem deserves further study.
bU
6
usually governs a direct object (patient).
til This sign has several meanings in Sumerian. In its reading as til, it is equated with
Akkadian gamaru, laqatu, and qatu. The CAD glosses qatu as " 1 . to come to an end, to be
used up, 2. to perish, 3. to become completed, fnished, settled". In the causative ster,
uqtu is glossed as "to bri ng to an end".
Lesson 15
Text 1 5
Column
,=
''
`

9
l
1 ! 4
+=

Column (continued)
1 The sign is partly restored from the
dplicate, No. V00JV.
167
168
Lesson 15
Column


> 1'|

"

==
^ =
Column (continued)

'


` "
*
|
Lesson 15 1
69
Notes
11 9: The nam-sign and the U1-sign are either partly effaced or poorly drawn on the
original. Not all scholars, in fact, believe that there even is a U1-sign in this particular text.
But since this curse-formula occurs in several other texts, its restoration is relatively certain.
I 1 :
2:
3:
4:
5:
6:
7:
8:
9:
10:
1 1 :
12:
13:
II 1 :
2:
3:
4:
5:
6:
7:
8:
9:
1 0:
1 1 :
Transliteation
d
Amar-
d
Zuen
Nibru
kC
1
d
En-lil-Ie
mu-pad-da
sag-us
_
d
En-lil-ka
lugil-kalag-g1
I I V
' ki
..w
-
nm
S
-ma
lugal-an-ub-da-liruzba-me
alam-ba
d
A
d
Z k' - V
ki
mar- uen
--
-1g!T nms -ma
mu-bi-im
alam-ba
lu ki-g.-ba-bi
i b-da-ab-kur-re-!
barag-sig
9
-g1-bi
i -bU
6
-re-!
d
Nanna
lugl-Vrims
kt
ma-ke
4
d
Nin-gil
V
ki
k ama- nm
S
-ma-

4
nam-U1-ba-an-da-kur
,
ne
numun-na-ni
h-eb-til-Ie-ne
Translation
I 1 : I am Amar-Sin,
2-4: proposed in Nippur by Enlil,
Transcription
Amarzuen
Nibru.a
Enlil.e
mu.pad.a
sagus
e.Enlil.(a)k.a(k)
lugal.kalaga
lugal.Vrim.a(k)
lugal.anub.da.(k) limmu.bLa(k).me.( en)
alam. bLa(k)
Amarzuen kLaga.a. Vrim.a(k)
mu.bLm
alam. bLa(k)
lu kLgub.a.bL0
Lb.da.b.kur.e.0.a
baragsiga. bL 0
L(b. )bur.e.0.a. (d)
Nanna
lugal.Vrim.ak.e
Ningal
ama. V rim.ak.e
namue.ba.n.da.kur. (e.e)ne.0
numun.ani.0
ue. (L)b.til.e.ene.0
170 Lesson 15
5-6: patron of the temple of Enlil,
7 the mighty king,
S the king of Ur,
9: the king of the four quarters.
1 0, 1 2: The name of this statue is:
1 1 : "Amar-Sin is the beloved of Ur".
11 5-6: May Nanna, the king of Ur,
7-8: and Ningal, the mother of Ur,
9: curse
1 3-2: the man who changes the place of this statue
3: and the man who tears down its pedestal,
1 0- 1 1 : and may they put an end to his offspring!
Commentay
1. The frst nine lines of this inscription are the same as in Text 1 3, except for the later
form of the royal title in line 7.
9. me is the frst person singular enclitic copula, "I am". In older Sumerian, it is usually
written -me-en. The form of the second person singular is also -me-en. The third person
singular is -am. Thus, the paradigm for the singular of the enclitic copula is:
frst person singular
second
third
-me-(en)
-me-@)
-am (after a consonant)
-m (after a vowel)
The distibution of -me - -me-en does not exactly correspond to a difference in time.
In older Sumerian, -me-en is the norm, but in Ur Ill, both -me and -me-en occur, with -me
predominating. However, both also occur in later texts. It is not sure if this should be
regarded as an orthographic or as a phonological problem; it is discussed further below.
The regular form of the enclitic copula for the third person singular is laml, written
with the am-sign. After a vowel , the enclitic copula appears as Iml, as in line 1 0 of Text
1 5.
10. The next three lines give the actual name of the statue. The construction is an
anticipatory genitive, with an enclitic copula: "of this statue, its name is . . . ". Thus, the bi of
mu-bi resumes alam: alam.bi. a(k) . . . mu.bi. m.
1 1 . The actual name of the statue is: "Amar-Sin is the beloved of Ur". This is an
equational sentence, and so one might have expected to fnd an enclitic copula. However, it
is not uncommon to fnd simple equational sentences without a copula, and this name may
be such an instance. It is also possible that the name is not a complete sentence, but rather
is a noun phrase with an appositive: "Amar-Sin, the beloved of Ur"; this is discussed
Lesson 15 171
further below.
The next few lines are rather complicated. Lines I: 1 3 through I: 4 are all the
comitative complement of the verb in I: 9. This complement includes a relative marker il
governing two relative clauses. The frst is marked by the - at the end of line IT: 2. The
second is marked by the -f at the end of line IT: 4. All of this is embedded inside an
anticipatory genitive:
"of this statue, the man who changes its position and tears down its pedestal" )
"the man who changes the position of this statue and tears down its pedestal".
The anticipatory genitive in I: 1 3 is resumed by the -bi in I: 1 and 11: 3.
11 - 1. ki-gub-ba ki.gub.a, "standing place" or something similar. Presumably, gub-ba
is a passive paticiple in .a.
2. Since the sense being conveyed is future, the verb is put into the maru. kur is a member
of the affxation class, so forms its maru with the maru-suffx .e, hence kur.{, written kur
re.
The initial lil of the i b-sign represents the conjugation-prefx i.
There are two Ibis in this paticular prefx chain; it is easier to look at the second Ibl
frst. Since this is a maru form of the verb, the Ibl in the slot immediately preceding the
verbal root cross-references the direct object , kiguba.bi.0.
The da is the dimensional-prefx which cross-references a nominal phrase in the
comitative case. The Ibl before the da is an element not yet seen. Before the dimensional
prefxes .da (cross-referencing the comitative .da), .i (cross-referencing the terminative
.e), and . ta (cross-referencing the ablative), it is possible for an "optional pronominal
prefx" to appear. For the third person, these prefxes are: .n for the animate, and .b for the
inanimate (forms for frst and second person are discussed below). These prefxes help
cross-reference the nominal phrases occuring in the sentence; they do not convey any new
information. Here, the Ibl refers back to alam of line 10. The use of these pronominal
prefxes appears to be purely optional (at least, no-one has fgured out any rules for thei
distribution); they did not appear in any of the previous texts used in this book, and only
show up sporadically in the remaining texts. This is not apparently a prblem in
orthography. That is, unlike the personal-afxes appearing immediately before the root,
one should not assume that these pronominal-prefxes were always present.
Thus, .bda represents the comitative dimensional-prefx, with an optional pronominal
prefx.
An obvious problem here is that there is no comitative nominal phrase in the sentence
for the dimensional-prefx to cross-reference. In fact, in this verbal prefx chain there oc
curs both a dimensional-prefx (.bda) and a personal-affx (.b). However, there is only one
noun phrase that these could cross-reference, ki-g!-ba-bi. Furthermore, this particular
verb seems to normally be construed with a direct object. Therefore, the personal-afx .b
cross-references the direct object, and the dimensional-prefx .bda does not refer back to
any particular nominal phrase.
As was mentioned in Lesson 1 2, there are numerous cases where a comitative
dimensional-prefx appears in the verbal prefx chain, but with no corresponding comitative
nominal phrase in the sentence. There is probably not just one single rule govering th(
17
2
Lesson 15
appearance or non-appearance of the comitative dimensional-prefx. More likely, there are
several different factors at work, which have not yet been unraveled.
To summarize the verb form:
i . b . da
( 1 ) (2) (3)
b . kur . e . 0
(4) (5) (6) (7)
( 1 ) conjugation-prefx
(2) optional pronominal-prefx
(3) comitative dimensional-prefx
a
(8)
( 4) personal-affx cross-referencing patient
(5) verbal root
(6) maru-sufx
(7) personal-afx crss-referencing agent
(8) nominalizer
The entire clause, nominalized in -!, stands in apposition to il forming a relative
clause.
4
. i-bu6-re-! L(b). bur.e.a. i is the conjugation-prefx. However, the verb has no dimen
sional-prefx, nor (on the surface) does it have any personal-afx cross-referencing the
apparent direct object, baragsiga.bi.0. Since the verb form is in the maru, one might have
expected to fnd /b/ immediately preceding the verbal root. There are at least two possible
reasons for its apparent absence. Lines 3-4 are roughly parallel in form and content to lines
1 -2. It is possible that its use in the verb form in line 4 would have been felt as redundant;
that is, the parallelism in construction peritted deletion of the personal-afx .b. Some
scholars have suggested that personal-affxes are only present (and therefore only written)
in ambiguous contexts.
However, there may be an entirely different reason for its absence. Perhaps there was
a phonetc reduction of /ibburea/ ) /iburea/. That is, the problem may be phonological, not
morphological. As is often the case in Sumerian, it can be difcult to determine whether a
problem is orthographic, phonological, or morphological in nature.
4. The assumed comitative case-marker .(d) at the end of this line marks lines I: 1 3
through I: 4 as the comitatve complement of the verb nam . . . kur
S
i n line 11: 9. The same
use of the assumed comitative with this verb occurred in Text 1 2.
6-8. The two agents of the verb fors in lines 9 and 1 1 are both marked by the ergative
case-marker .e.
9 . . nda is the comitative dimensional-prefx plus the optional pronominal-prefix . n. Here
the animate form is used, because it refers back (essentially) to "the man who . . . ".
All the verb forms seen up to this point have been singular. Here, there is a plural
agent ("Nanna and Ningal"), and so the verb must be put into the plural (Sumerian has no
dual). The plural third person of a maru verb is usually written with a suffxed --ne.
Sumerologists have different understandings about the morphology implied by this writing.
One analysis sees this writing as refecting: e.ene.0. The .e is the mat-suffix; .ene is the
plural marker; and .0 is the personal-afx cross-referencing the agent. Another analysis is
Lesson 15 173
to isolate the morphemes of this ending as: e.0.ene. A third analysis is to read the ne-sign
as de, refecting quite a different understanding of the morphology. This is a very difcult
issue to resolve. Here, the frst analysis has been followed.
A problem in this particular verbal form is the fact that only -ne is written, not --ne.
How should this writing be understood?
The Falkenstein school would read the frst sign with an overhanging vowel, that is,
kure2' A variant of this solution is to read the kur-sign as kUfU
S
: kUfU
s
-ne would represent
kur.e.ene.0. The second lul in the sign kUfU
S
would represent the assimilation of the un
derlying lel to the frst lu/. There are numerous other cases where the mari-suffx .e is
assimilated to an I ul -vowel of a verbal root.
But has been stated several times, the mnemonic nature of the script may have meant
that there was no need for the scribes to wrte down the full forms of the morphemes.
To summarze the verb form:
nam be ba n da kur . e . ene 0
( 1 ) (2) (3) (4) (5) (6) (7) (8) (9)
( 1 ) nominal component of compound verb
(2) modal-prefx
(3) conjugation-prefx
(4) optional pronominal-prefx
(5) comi tati ve dimensional-prefx
(6) verbal root
(7) mara-suffx
(8) pI ural marker
(9) personal-afx cross-referencing agent
This line is essentially the same curse-formula which occurred in Text 1 2. It is
instructive to compare the verb forms of the two texts:
Text 1 2 nam-bf-ba-da-kur
s
-
Text 1 5 nam-bf-ba-an-da-kur
s
-ne
The frst difference is the presence in Text 1 5 of the animate optional pronominal
prefx .n before the comitatve dimensional-prefx. It is precisely cases such as this - two
texts with a minimum of varation - which seem to show that such pronominal prefxes are
indeed optional.
The second difference is in the number of the verb. In Text 12, the agent of the verb is
singular ("Gilgamesh"); in Text 1 5, the agent of the verb is plural ("Nanna and Ningal").
1 1 . b is the regular form of the desiderative modal-prefx. In its two previous
occurrences, Ibe/ ) Ibal before the conjugation-prefx Iba/.
The next slot in the verbal prefx chain should be the obligatory conjugation-prefx. I
this particular case, the conjugation-prefx 1 has contracted into the I el of the modal-prefx .
. b cross-references the direct object, numun.ani.0. Since numun refers to "future
descendants, progeny", it might seem a little surprising to see numun. ani.0 cross
referenced by .b (normally used for inanimates) instead of by .n (normally used for
animates). However, .b is frequently used for what might be considered "collectives" (both
1 74 Lesson 15
of animates and inanimates), and here numun was probably felt as a collective.
Since til is a transitive verb, b is used with the maru form of the rot. til is a member
of the affxation class, so its maru-root is t1.e. The verb is written as expected, til-Ie-ne.
To summarize the verb form:
b
e
( 1 )
i
(2)
b
(3)
( 1 ) modal-prefx
tl
(4)
(2) conjugation-prefx
e ene
(5) (6)
(3) personal-affx cross-referencing patient
(4) verbal root
(5) maru-suffix
(6) plural marker
(7) personal-affix crss-referencing agent
Discussion: structure
This text is composed of three sentences:
I: 1 -9
I: 10- 1 2
I : 1 3-11: 1 1
nominal sentence
nominal sentence
verbal sentence
The stucture of the verbal sentence, in essence, is:
I: 13-II: 4
11: 5-6
11: 7-8
11: 9
11: 1 0
11: 1 1
accompaniment
agent
1
agent2
verb
patient
verb
o
(7)
In a famous monograph entitled Das apositionell bestimmte Pronomen ( 1 932), Poe
bel tried to show that the copula in Semitic and in Sumerian could also be used to express
apposition. In Text 15, the translation of the beginning of the inscription would be: "I,
Amar-Sin, the one nominated ... ". That is, the frst nine lines would not form a complete
sentence, but rather would form a kind of casus pendens or fronting for emphasis. Simi
larly, the name of the statue in line 1 1 may be a noun and appositive, and not an equational
sentence.
- Sign formation
In Lesson 3, it was pointed out that some cuneiform signs are in origin combinations
of a pictographic sign with an inscribed phonetic indicator of some kind. The ama-sign is
thought to be one of these. It has an inscribed digir-sign, one of whose phonetic readings
is am
6
. Therefore, it has been suggested that this component of the sign is a clue to the
Lesson 15 175
pronunciation. Unfortunately, it is diffcult to make this square with the fact that the rest of
the sign appears to be the pisan-sign, which basically means "box" of some kind. It has
also been proposed that there is some obscure symbolism involved, "mother" being
represented as a "divinity" within a "box"!
-Loan words
In Text 6, the word barag occurred; this term was borrowed into Akkadian as parakku.
The amissable /g/-Auslaut shows up in Akkadian as a geminated voiceless consonant.
barag-sig
9
-gi, however, was borrowed into Akkadian as barasigu. The amissable / g/
Auslaut of the barag element, here in syllable-fnal position, does not appear. However, the
intervocalic / g/ of g-gi (sig.a) remains. The word-initial /b/ of barag is teated dif
ferently in each loan word. In parakku it is refected as /p/, but in barasigf it is refected as
/b/. This shows that barasigu is a later borrowing into Akkadian than parakku. It is us
ually assumed that in relatively older loan words from Sumerian, Sumerian voiced stops are
refected in Akkadian as voiceless stops. In relatively later loan words, the same Sumerian
voiced stops are refected as voiced stops. Needless to say, not enough is known about the
phonetics of either Sumerian or Akkadian to explain exactly what has happened. However,
such differences are one way that the entry of loan words into another language can be
dated relative to each other.
-Conjugation -prefxes
In the model of the Sumerian prefx chain used throughout this book, the use of the
conjugation-prefx is obligatory; a conjugation-prefx is present in every fnite verbal form.
Therefore, in line 11: 1 of Text 1 5, it is assumed that a conjugation-prefx 1 has assimilated
into the modal-prefx b. Similarly, in Text 22, this view assumes the presence of a
conjugation-prefx 1 after the cohortative modal-prefx g.!, although the verb form in Text
22 is written g.! -an-
!
-
i
.
Although both the desiderative modal-prefx b and the cohortative modal-prefx g.!
are fequently followed by such conjugation-prefxes as mu and ba, there appear to be no
instances of writings of the type *b-l or *g.! -l. If the 1 is in fact present, one might expect
to fnd at least a few occurrences of it being written (to judge by similar phenomena in the
script). However, none apparently occur. This means that the assumption that the attested
spellings all represent assimilation may not, in fact, be correct.
Jacobsen, for example, believes that the conjugation-prefx 1 is "incompatible" with the
cohortative modal-prefx g.! . That is, the semantic information conveyed by 1 does not
permit it to co-occur with g.!. This may mean that at times the modal-prefx g.!is followed
by no conjugation-prefx. The up-shot is that not every fnite verbal form contains a
conjugation-prefx.
This problem cannot be resolved here. However, it should be kept in mind that the
general principle stated in this book -that conjugation-prefxes are obligatory - may need
modifcation, particularly in the case of certain modal-prefxes.
176
Lesson 15
-Pronominal-prefxes
The use of pronominal-prefxes before certain dimensional-prefxes is not uncommon.
In the singular, the basic fonns of these prefxes appear to be:
frst person singular . 0
second .e
third animate . n
inanimate . b
The form of the frst person singular was probably not simply .0. Some irregular
writings have led scholars to speculate that the frst person singular marker had either a
vocalic component (perhaps lel, similar to that of the second person), or a consonantal
component (perhaps a semi-vowel or glottal consonant). The fonns of the plural are much
less clear.
If more than one dimensional-prefx occurs in a verbal prefx-chain, it appears that
only the frst dimensional-prefx can have an optional pronominal-prefx.
An older view of these pronominal-prefxes was that they were obligatory, and that
their relative infrequency in texts is due to vagaries of orthography. In this view, for
example, every .da dimensional-prefx which cross-references a third person should be un
derstood as: (n.)da. Thus, the verb fonn in Text 1 2, nam-b-ba-da-kur
5
-, should be un
derstood as: nam.be. ba.(n. )da.kur.e.0, instead of: nam.he.ba.da.kur.e.0. However, this
view is not as widely accepted today, because the actual number of cases where an .n or .b
appears before an appropriate dimensional-prefx is much less than the number of cases
where they do not appear.
-Dimensional-prefxes
In line 2, the verb kur was used with the dimensional-prefx -da, although the sentence
contains no nominal phrase in the comitative case. In some cases, verbs have become
lexicalized with certain dimensional-prefxes; that is, the verb will frequently (sometimes
always) have a certain dimensional-prefx, even if no corresponding nominal phrase occurs.
Occasionally a noun phrase will be marked with a certain case-ending, but the
dimensional-prefx used in the corresponding verbal phrase will be different than e,pected.
For example, a nominal phrase in -e (the tenninative) may be resumed by the dimensional
prefx nonnally used for the locative (-ni-). In some cases, this may have happened
because of a historical change in the rection of a noun phrase. That is, at one time a noun
phrase may have been marked by one particular case, but in time the case which was used
changed. However, the dimensional-prefx, being more closely bound, did not change.
This can lead to such cases as a teninative being cross-referenced by a locative. In the
case of kur, for example, its complement may originally have been in the comitative case (or
perhaps the ablative case), but it eventually shifted to the absolute case. However, it caried
along its dimensional-prefx, resulting in such cases as line 11: 2, where the verbal prefx
chain has both a dimensional-prefx and a personal-affx, yet there is only one nominal
Lesson 15 177
phrase for both of them to govern.
Similarly, the rection of the frst element of a compound verb may change. In earlier
texts, it may be in one of the adverbial cases; in later texts, it may be constued as a patient.
Sometimes this change may be due to Akkadian infuence. Unfortunately, there is rarely
enough data to prove such hypotheses for any paricular verb.
The analysis of 11: 1 -2 presented here is not the only one possible. 11: 1 was explained
as the patient of the verb, and the da in the verbal prefx chain as a frozen use of the
comitative dimensional-prefx. However, it is known that the verb kur sometimes takes its
complement in the ablative case. Therefore, it has been proposed to understand line 1 as:
kLgub.a.bi. (ta), with the ablative case-marker not expressed in writing. The dimensional
prefx which cross-references the ablative is Ita/, normally written ta. Therefore, the verb
form in line 2 would have to be understood as: Lb.ta.b.kur.e.0.a. The fact that the text
clearly shows a da-sign, and not a ta-sign, would seem to mitigate against such an
interpretaton. However, other cases of the ablatve dimensional-prefx appearing on the
surface as da instead of ta are known; the reasons why are unclear. In our particular text,
there could have been assimilation of voice: Ibta/ ) Ibda/. However, other such cases of
an apparent da for ta are less amenable to phonetic explanations of this sort.
-Enclitic copula
The fact that the frst and second persons of the enclitic copula occur both as -me and -
me-en admits of several possible interpretations. The problem may have been orthographic:
word and syllable-fnal nasals often seem not to be written in Sumerian (cf. mu-na-du for
mu.na.(n.)du). It may refect the phonetic process of Sumerian dropping word-fnal nasals;
in this case, the forms written with -en are to be regarded simply as morphographemic or
archaic writings. It has also been posited that the writings -me and -me-en represent one
pronunciation, /me/. The original word-fnal nasal was lost, producing a nasalized vowel,
which the script could not well represent.
-Proper names
Votive objects of all kinds were often given names. This has been studied by Gelb,
who points out that such names are often complicated sentences in their own right. For
example, a statue of Gudea dedicated to Ningirsu was given the name (Gelb
'
s tanslation):
"N ingirsu, -the-king( =god)-w hose-heavy-might -the-word-cannot -bear, -has-decided-good
destiny-for-Gudea, -who-built-this-temple" (1956:66).
- History
In addition to the copy ofthe text reprduced above, a late, Neo-Babylonian copy frm
the seventh century BC is also preserved. It was inscribed on what was apparently a model
pedestal. This copy is interesting because of the presence of several errors in the Sumerian.
Also, it has a colophon written in Akkadian, which seems to say that the model was to be
used in an "exhibition" (tamaru) of some kind. The page following is a photograph of two
sides of this Neo-Babylonian version of the inscription.
The following quotation from Woolley describes this object, and also says something
178 Lesson 15
about the tenor of the times which produced it:
A little way apart lay a small drum-shaped clay object on which were four
columns of writing; the frst three columns were in the old Sumerian
language, and the contents of one at least were familiar to us, for we had
found it on bricks of Amar-Sin, king of Ur 2046-2038 B. C., and the other
two were fairly similar; the fourh column was in the late Semitic speech.
"These", it said, "are copies from bricks found in the ruins of Ur, the work
of Amar-Sin king of Ur, which while searching for the ground-plan [of the
temple] the Goveror of Ur found, and I saw and wrote out for the marvel of
beholders". The scribe, alas! , was not so leaed as he wished to appear, for
his copies are so full of blunders as to be almost unintelligible, but he had
doubtless done his best, and he certainly had given us the explanation we
wanted. The room was a museum of local antiquities maintained by the
princess Ennigaldi-Nanna (who in this took after her father, a keen
antiquarian), and in the collection was this clay drum, the earliest museum
label known, drawn up a hundred years before and kept, presumably
together with the original bricks, as a record of the first scientific excavations
at Ur . . . . We shall see furher examples of the archaeological spirit that
prevailed in the latter days of Babylon, but undoubtedly it was reinforced by
a pathetic superstition that looked back across the uncounted ages to the
fabulous beginnings of things when men and gods were scarcely to be
distinguished and "there were giants in the land in those days" (1982: 252,
231 ).
L0880n1
I79
Lesson 1 6
This is another door socket of Amar-Sin.
Sign-list ad vocabular
_ h honey
i oil, fat
' nun prince, noble
_ i-nun butter
Notes
siskur2
. .
vme, wme
sacrifce
Wig to cease
! and
nu not
1 It is thought that the bee is native to Syria, but not to Mesopotamia. So Il (and its
Akkadian equivalent dgpu) may be a Syrian import. It has also been speculated that h and
dispu were actually a kind of syrup made from fuits.
i-nun Literally, "oil of the prince". Some scholars interpret this as "butter", others as
"ghee" (that is, clarifed butter). The latest discussion (K. Butz, 1973-74:37) interprets it as
"Butterschmalz" ("butterfat"). The Akkadian equivalent of i-nun is bimetu; this is tans
lated by the CAD as "ghee".
On the surface, i-nun would appear to b a noun-noun compound, of the type dis
cussed in Lesson 9. However, there are spellings in other texts which show that this is a
genitive phrase, "oil of the prince", Lnun.a(k). For example, in Gudea this expression is
written both i-nun and i-nun-na. The term i-dug-nun-na, i.dug.nun.a(k), "good oil of the
prince", also occurs, but it is not sure what this means.
In older works, the word for "oil, fat" is transliterated as hi, instead of i.
gdtin The word for "wine" also appears as tin. gestin, therefore, may originally have
1 1
1
8
2
Lesson 16
meant "wine-tee, vine, grape", etc.; the meaning then became extended to "wine". The
Akkadian equivalent, karanu, is glossed by the CAD as: "1 . wine, 2. grapewine, 3.
grapes".
sisku2 This word is normally written just as it appears here, by graphic reduplication of
one sign. This sign appears to be the amar-sign inside of which is the barley-sign, Se.
Thus, the original, pictorial signifcance of the sign may have been "grain-fed cattle", or
something similar.
Sometimes the word Isiskurl is written with only one, instead of two, signs. In this
case it is tansliterated as siskur. Earlier Sumerologists thought that the writing with one
sign was a singular, and the writing with two signs was a plural; this means that the writing
with two signs was sometimes transliterated as siskur-siskur. The word is further dis
cussed in Lesson 1 9.
The prnunciation of the sibilants is unsure, and so it is also transliterated as siskur2'
sizkur2
'
Siskur2, etc.
ilig O, silig. This root is very uncommon in verbal forms; it is mostly used in participial
orinfnitival constructons.
Lesson 16
1
0
XI

1 83
18
4
1 :
2:
3:
4:
5:
6:
7:
8:
9:
1 0:
1 1 :
12:
1 3:
14:
1 5:
1 6:
17:
1 :
2:
3:
4:
5:
8:
10:
1 1 :
1 2:
17:
13:
14:
1 5:
Transliteration
d
En-lil
lugil-kur-kur-ra
lugil-ki-.M-M-ni-i
d
Amar-
d
Zuen
d
En-lil-Ie
Nibru
kC

mu-pad-da
sag-us
_
d
En-lil-ka
lugal-kalag-g
I I U
'
ki
..g- nms -ma
lugil-an-ub-da-limmurba-ke4
lal i-nun
y getin
ki-siskur
T
ra-ka-na
nu-i1ig-g
mu-na-an-du
Translation
For Enlil,
king of the lands,
his beloved king -
Amar-Sin,
proposed by Enlil in Nippur,
patron of the temple of Enlil,
the mighty king,
the king ofUr,
the king of the four quarters -
built -
the temple (where) honey, butter
and wine
in his place of sacrifce
Lesson 16
Transcripton
Enlil
luga1.kur.kur.a(k)
luga1.kL aga.a.anLr
Amarzuen
Enli1.e
Nibru.a
mu.pad.a
sagus
e.Enli1.(a)k.a(k)
luga1.kalaga
luga1.Urim.a(k)
luga1.anub.da. (k) limmu.bLak.e
e lal inun
u getin
kLsiskur.ak.anLa
nu.i1ig.e(d).0.0
mu.na.n.du.0
1 6: shall not cease.
Commenta
Lesson 16
12. The .e marks the agent, which spans lines 4 through 12.
185
13-16. These lines are the direct object (patient) of the verb in line 17; they are marked as
such by the fnal . 0 at the end of line 1 6.
means "temple"; the rest of the text is apparently a relative clause modifying. Up to
this point, all the relative clauses which have occurred have been verbal sentences. Here,
however, the relative clause is a nominal, equational sentence: "honey, butter and wine in
his place of sacrifce are a non-ceasing thing". The syntax of nominal relative clauses is
little understood. They consist of a nominal element as subject and a nominal element as
predicate; the enclitic copula is not used. Unlike the case with verbal relative clauses, no
nominalizer is used. Also, no relative marker is used. This entire equational sentence is in
appositon to the noun .
Although Sumerian does not use any relative marker, English needs a relative adverb
of some kind, such as "where".
14. Sumerian does not normally use any conjunction between nouns; instead, it prefers to
conjoin them directly. However, Sumerian also has an option of using the conjunction 1
between nouns or nominal phrases (rarely between verbs). Here, it is used between the
second and third nouns of a list: "honey, butter and wine".
1 is not a native Sumerian word; it is borrowed from Akkadian. It is not common for
languages to borrow such syntactic devices as conjunctions. This is an indication of the
Sumerian language giving up some ground before the onslaught of Akkadian. 1 turns up
as early as the Tell Abu Salabikh tablets (2600 BC). According to the editor of these
tablets, Robert Biggs, "even at this early date Sumerian may have been under a heavy
Semitic infuence" (1974: 32). (Post Ur III texts also occasionally use the Akkadian
conjunction -ma.)
15. The fnal .a is the locative case-marker. It is not resumed by any dimensional-prefx.
16. nu is the general, all-purpose negative marker in Sumerian. It can appear before verbal
forms and before nominal forms.
When used with verbal forms, nu falls into the category of modal-prefx. Thus, it
never co-occurs with any other modal-prefx. It is followed instead by a conjugation
prefx. (The conjugation-prefx 1 often assimilates into the lul of nu.)
ilig is apparently an active participle from the verbal root "to cease" .
. ed is an element not seen before. It is one of the most puzzling of all Sumerian
morphemes. It normally follows manl-forms of the verb; its use with bam!!-forms is very
rare. It is also used with both active and passive participles. With active participles, it
occurs as: du.ed.0. With passive participles (much rarer), it occurs as: du.ed.a. The form
in Text 1 6 is: ilig.ed.0.
The meaning of .ed has been much discussed. It must have something to do with the
"future", although the exact nuance being conveyed is unsure. In its very rare occurrences
with the bar!!, it appears to have a future-perfect sense. According to Jeremy Black,
Its reference seems to be to future events, although its use in descriptive
1 86 Lesson 16
passages suggests a connotation of vividness. More important, it marks an
action as not yet begun at the moment of obseration (which can be a
moment in the past) ( 1 984: 1 1 8).
According to Jacobsen, .ed is the
mark of pre-actional aspect indicating prospectiveness of the action as
present at the point in time the speaker has in mind. Attention is thus not on
the action as future but on its prospectveness as present ( 1965:267).
Because Idl is an amissable consonant, it often does not show up in the script. This
can make it very difcult to recognize.
Line 16 forms the predicate of the nominal relative sentence. This predicate is a
nominal form: "something which will not cease in the future", i1ig.ed.0. Thus, the frst
.0 in transcription marks ilig.ed as an active participle. The second .0 marks the entire
complex as the patient of the verb in line 17.
At frst glance, it might seem possible to interpret nu-ilig-g as a verbal, not a nominal
form. However, this analysis would seem to be precluded, because of the lack of a final -.!
in line 1 6. If it were a verbal form, it would need to be nominalized in -.! to serve as a
relative clause.
17. mu-na-an-du mu.na.n.du. This is the frst time in this body of texts where the ani
mate personal-affx .n, cross-referencing the agent in the bamn!, appears explicitly in the
script.
There is no dimensional-prefix crss-referencing the locative nominal phrase in line
15.
Discussion: structure
The structure of this text is:
[Enlil, lugal.kur.kur. a(k), lugal.kLaga. a.ani].r
[Amarzuen, Enlil.e Nibru.a mu.pad.a,
sagus.e.Enlil.(a)k.a(k), lugal.kalaga,
lugal. Urim.a(k),
I ugal.anu b.da. (k)limmu.bLak].e
[e lal inun u getin kLsiskur.ak.anLa
nu.ilig.e(d).0] .0
mu.na.n.du.0
benefactive
agent
patient
verb
Relative clauses consisting of a nominal sentence, such as lines 1 3- 1 6, are not
common in Sumerian, and so there is some queston about the analysis. It is possible that
these lines should be understood as the name of the temple: "Amar-Sin built a temple
(whose name is) 'honey, butter and wine in his place of sacrifice shall not cease"
'
. In such
constructions, Sumerian does not always use the word for "name" (mu); instead, it conjoins
the name directly, as an appositive. An example occurs in Text 19: bad-Mar-tu Mu-ri-iq
Ti-id-ni-im, "the Martu-wall (whose name is) 'Murq-Tidnim"
'
.
Lesson 16 187
- .ed
There is no standard ter used to refer to .ed. It can appear both on fnite verbal
forms, and on nominal forms derived from verbal roots. On fnite verbal forms, it occurs
with both mari-roots and bamt!-rots (although there has been much argument about this).
It seems to ocur with both tansitive and intansitive-passive forms. On nominal forms, it
occurs on the "infnitive", the active participle, and the passive participle.
Because the Idl is amissable, and because the lel can assimilate into other vowels, it
is not always easy to determine even if it is present or not. The assumed Id! does not show
up in Text 1 6, and it is therefore also possible that some other interpretation of line 1 6 is to
be prefered. In spite of several recent attempts to unlock its morphology and syntax (the
most recent is that of Gerd Steiner [ 1 980]), it remains elusive.
-Participles
The taditional view of Sumerian is that it has two participles, an active and a passive
intransitive (Poebel used the terms nomen igentis and nomen actionis). The active ends in
.0, and the passive-intransitive in .a. Each participle can also appear with the element .ed.
This yields four possibilities:
X.0
X.ed.0
X.a
X.ed.a
However, a fair number of exceptions seem to occur. There are cases of participles in
.0 which seem to have a passive meaning, and participles in .a which seem to have an
active meaning. For example, the form in Text 1 6 is mig-g, presumably for ilig.ed.0.
Here it is being used intransitively, and therefore one might have expected a participle in .a.
The situation was undoubtedly more complex than the above table would indicate. Most
discussion of non-verbal forms in Sumerian has applied Indo-European or Semitic
grammatical categories and terms to the S umerian forms.
The most recent, full-scale, study of the participles is that of I.T. Kaneva ( 1 970). She
analyzes the forms differently:
X.0 "transitive participle of the imperfect aspect"
X.a "transitive participle of the perfective aspect"
X.a "intransitive participle"
However, she does not apparently recognize the existence of reduced relative clauses,
whose existence complicates the picture.
-Personal-afxes
This text marks the frst time that the personal-affix .n cross-referencing the bamt!
agent appears in the writing in the body of texts studied here. As discussed in Lesson 14,
this is part of the on-going process of the Sumerian writing system becoming more and
more explicit.
188 Lesson 16
-Conjunctions
The native Sumerian conjunction meaning "and", linking nouns and nominal phrases,
is -bi-da, sufxed to the second noun: an-ki-bi-da, "heaven and earth". It is thought that bi
here is the possessive-suffx, and da is the comitative case-marker. Thus, this originally
meant "the heavens along with its earth".
Sumerian also has a conjunction meaning "and", linking verbs: -in-g, ocurring on
the last verb of a series of two or more, in the position after the modal-prefxes but before
the conjugation-prefxes. No examples seem to ocur in the Ur III royal inscriptions. A
late example is: . . . za- in-g--zu, "you also know", for: inga.i .e.zu. Because in-g is
used so infrequently, it may have had some emphatic value, instead of straight co
ordination.
- History
This door socket (along with another, having a different inscription) was found in
what has been described as the "small shrine" in Nippur, part of a large temple complex
dedicated to Enlil. The door socket bears traces of an inscription of Lugal-kiginne-dudu,
who ruled in Uruk approximately 2400 BC (that is, some three hundred years before
Amar-Sin); this early inscription was also dedicated to Enlil. Thus, Amar-Sin
'
s builders
did not fashion a new door socket - rather, they engraved their inscription upon this
already-extant door socket, which presumably they had uncovered in their work of
rebuilding the Enlil temple.
-Literary parallels
The combination "honey, butter and wine" occurs in other Surerian texts. Sometimes
the three elements are listed in the same order as in this inscription; sometimes the order
varies. For example, in a literary text entitled "Nanna-Sin's Journey to Nippur", there
occurs: "May Nanna-Sin make butter, honey, and wine (i-nun lal geStin) abundant"; this
particular line is repeated verbatim fve times.
The expression i-nun has recently been found in one of the Sumerian-Eblaite voca
bulary lists. Unfortunately, only the Sumerian for this particular expression i s given, not its
Eblaite equivalent. One of the other entries in this vocabulary list, however, is for "good
oil". The Sumerian column is i-dug; the Eblaite equivalent is sa-ra-nu t-bu.
Lesson 16 189
Text 1 6a
supplementary
This inscription was engraved on a small bead of agate, used as a votive object.
Because of its small size and somewhat irregular surface, the signs are less than elegant.
No photograph is available.

` \`
'f Y
\.
/ (
/ ', j ;;<->/
1 90 Lesson 16
Notes
Commentay
8. . serves as a conjunction between two nominal phrases.
9. Nin-tur-dumu-g!l
O
. This appears to be a personal name, although there do not seem to
be any parallels to such a formation ("Nintur is my child" ?).
10. Both here and in line 7, the nominal phrases are marked by the terminative case-marker
-e.
16. It may be useful to compare the different forms of the standard curse-formula which
have occurred:
Text 1 2
1 6a
15
nam-b-ba-da-kur
5
-
nam-b-ba-da-kur5re
6
nam-b-ba-an-da-kur5ne (plural)
The difference between Text 1 2 and Text 1 6a is in the orthography of the maru-sufx.
In Text 1 2, a morphemic spelling is used. In Text 1 6a, the spelling follows the more usual
rules of Sumerian orthography.
Lesson 1 7
Tis is a large door socket of Amar-Sin.
1VJ
1
9
2 Lesson 1 7
Sign-list and vocabulary
T
Kar-zid-da Karzida (GN)
kar quay, pier; market-place
_ utu sun
kalam land
<r

g
-Il
giparu (part oftemple complex)

.
`

o
Notes
en priest, priestess
zid W) effective, true
ul remote, distant
l Gl
)
kur
9
(ku4)
sud
4 1 3)
li
ta
un
ab
bright, pure, holy
to enter
to be long
Kar-zid-da The name means "the effective quay", kar.zid.a. Karzida was apparently a
quay at Gaesh (written Ga-eS
ki
and Ga-eSs
ki
). The latter was close to Ur, and evidently
had some kind of cultic connection with it; it had its own Nanna-temple. Neither the site of
Karzida nor of Gaesh has been identifed.
kalam This word is apparently only used to refer to Sumer itself. It only occurs in the
singular, and is especially common with a possessive-suffx. The original pictographic
value of the sign is unknown. This word is further discussed in Lesson 1 9.
gi
6
-pA This was a secton of the temple complex; it was where the en-priest or en-
Lesson 1 7
1
93
priestess lived. It was borrowed into Akkadian as giparu. It is discussed further below.
en Besides its use in political contexts (where the conventional translation is "lord"), en
can also be used to refer to a specifc kind of priest or priestess.
zid It is diffcult to pin down a precise meaning for zid; it can mean something like
"someone (or something) that dos what he/it should be doing". The Akkadian equivalent
is kInu, glossed by the CA as: " 1 . tue, reliable, just, 2. honest, decent, loyal, 3. correct,
normal, regular, sound, legitimate".
ul This sometimes appears as ul, as in, for instance, uru-ul, "the primeval city". However,
sometmes an lit follows, resulting in graphic reduplicaton: ul-H. And sometimes ul or
ul-li is followed by la/, presumably the same nominalizer seen in such adjectives as kalag
g. In Text 17, it is wrtten: u
Hi
-.
It is probably related to ul meaning "bud" of a fower.
kur9 In older works, usually tansliterated as tu; some modern-day Sumerologists still
prefer this reading.
sud4 The verbal root meaning "to be long" can be written in two different ways in
Sumerian. The most common writing is: )--, read as sud (or su). However, it can also
be written (as in Text 17) as: >- . Older works tended to understand this last sign as
gid, but most scholars now read it as sud4 (or su
13
)'
It can be seen that the sud4 -sign is basically just a non-gumlfed version of the sud
sign. In general, it is always possible to use a non-gunufed sign in place of a gunufed
sign. For example, the gi-sign is frequently found in place of an expected g4. In Lesson
1 , it was mentioned that the name of the city of Ur is usually spelled Se-ab, but also Se
unug. It is possible that the unug-sign is a gunufed ab-sign.
This is why >-- is now read sud4 instead of gid. At one time, it was thought that
Igidl and Isud! were two different roots for "to be long", but now it is assumed that they
are just variant writings of the same rot I sud!.
The fnal consonant of the rot was probably the Idr/-phoneme discussed in Text 1 2.
In Text 17 it is followed by the re
6
-sign (with its initial Ir/).
Le
sso
n 1 7
194

0X

1
J
Lesson 1 7
1
95
Notes
1 1, 18. In this text, the kalam-sign and the un-sign ae very simila in shape. The histo
rical relationship between the two signs is complex. It is usually assumed that in origin
they were formally distinct. In caefully executed texts, they ae stll distinguished through
the Old Babylonian period. For example, in the Code of Hammurapi, the kalam-sign ap
peas as: , and the un-sign appears as: If- .
However, the un-sign (more properly, y; its value was either lugl or lugu/) means
"people", and there is a close semantc connection between "people" and "land". This
might mean that the signs were not formally distinct in origin, but perhaps became
secondarily differentiated.
By the Neo-Assyrian period, the two signs fall together into one sign. There ae
several Neo-Assyrian signs which represent the confation of two or more signs which in
origin were different signs. For example, the Neo-Assyrian ku-sign is a continuation of
several different box-shaped signs, which have all fallen together in shape.
17. The nu-sign is poorly drawn; a better drawn version appeas in line 1 8. There appeas
to be an extra horizontal stroke, running into the top of the du-sign.
1 :
2:
3:
4:
5:
6:
7:
8:
9:
10:
1 1 :
12:
1 3:
14:
1 5:
1 6:
17:
1 8:
Transliteration
d
Nanna
Ka-zid-da
lugal-ki-

g-g

-ni-!
d
Amar-
d
Zuen
d
En-lil-Ie
Nibru
kC
i
mu-pad-da
-us
_
d
En-li l-ka
il i-zid
d
Utu-kalam-ma-na
lugal-kalag-gi
I I U '
ki
..gf- nm
5
-ma
lugal-an-ub-da-limmuzba-ke4
Ka-zid-da-i
ud-ul-li-i-ta
-par-bi nu-du-am
en nu-un-til-Ia-am
Transcription
Nanna
Kazida.(k)
lugal.kLaga.a.ani.r
Amazuen
Enlil.e
Nibru.a
mu.pad.a
sagus
e.Enlil.(a)k.a(k)
digir.zid
Utu.kalam.anLa(k)
lugal.kalaga
lugal. Urim.a(k)
lugal.anub.da. (k) limmu.bLak.e
Kazida.a
ud.uli.a.ta
gipa.bi. 0 nu.(i .)du. 0.a.am
en.0 nu.(i .)n.ti1.0.a.am
1
96
19:
20:
21 :
22:
23:
24:
25:
26:
27:
28:
d
Amar-
d
Zuen
ki-g-
d
Nanna-ke4
-Ql -kug -g -ni
mu-na-du
en-ki-ag-ga-ni
--- - -
mu-na-ni-kur
9
d
Amar-
d
Zuen-ke4
ud im-da-ab-sud4-re6
nam-tl l-Ia-ni-e
-mu-na-ru
Translation
1 : For Nanna
2: of Karzida,
3: hi s beloved king -
4: Amar-Sin,
5-7: proposed by Enlil in Nippur,
8-9: patron of the temple ofEnlil,
10: the effective god,
1 1 : the sun-god of his land,
1 2: the mighty king,
1 3: the king of Ur,
1 4: the king of the four quarters -
1 5: in Karzida -
1 6:
from of old -
Lesson 1 7
Amarzuen
kLaga.a.N anna.k.e
gipar.kug.ani.0
mu.na.(n.)du.0
en.kLaga.a.anL0
mu.na.nL(n.)kur.0
Amarzuen.(a)k.e
ud.0 1 .b.da.b.sud.e.0
nam. tl.anUe
a.mu.na. (n.)ru
17: - its giparu not yet having been built -
1 8: - and no en having taken up residence in it -
19: Amar-Sin,
20: the beloved of Nanna -
22: built
21 : his pure giparu.
23-24: He made his beloved en-priestess enter it.
25-26: Ama-Sin will prolong (its) days.
27: For the sake of his long life
Lesson 1 7 197
28: he made a votive offering.
Commenta
1-2. Presumably this is a genitive phrase. A parallel constuction occurred in Text 1 2:
d
Bi I-g-me En-dim-gig
ki
il gal--ni, "for Gilgamesh of Endimgig, his king".
10- 1 1. These two epithets did not occur in the previous inscriptions of Amar-Sin. Amar
Sin is not just content with the use of the divine determinative before his name; he must
also refer to himself as "effective god" and as "sun-god of his land". (Shulgi is also
referred to as digI kalam-ma-na.)
14. The .e at the end is the ergative case-marker.
15. The writing in - is ambiguous. The original editor of this inscription interpreted it as
an anticipatory genitive: "of Karzida, its giparu" ) "the giparu of Karzida". However, it
may also represent a locative case-marker: "in Karzida". The . n in line 18 below would
seem to favor this latter interpretation.
16. Literally, "from a distant day", that is, something like "from of old", "since time
primeval". ta is the marker of the "ablative" case. This case can usually be translated as
"from", e.g. , il gal-ta, "fom the king". It is very common in the formation of adverbial
phrases.
The ablative is usually cross-referenced in the verbal prefx chain by the ablative
dimensional-prefx ta. However, in formulaic adverbial constructions such as this one, it is
not normall y cross-referenced.
This particular expression is not uncommon. Outside of Ur Ill, for example, it occurs
in a Gudea inscription: ud-ul-li-i-ta numun--i-ta, "fom of old, frm when seed (frst)
came forth" , "to go out", Akkadian wasil. ud-ul is also used in the formation of other
adverbial phrases, e.g., ud-ul-Ia-e ud.u1.a.M, "for a long time", "forever".
17. All the verb forms seen up to this point have been in what we would call in English
the "active voice". There has been a long discussion about whether or not Sumerian has a
"passive voice". Some Sumerologists say that Sumerian has no passive voice. Others say
that Sumerian is basically passival in nature.
To some degree this is a question of linguistc theory, and not of Sumerian. It is a
question of the defnition and nature of active and passive, of the contrast between passive
and intransitive, and of the way such distinctions are marked in the morphology and in the
syntax. The problem is exacerbated by a tendency in the past to transfer categories found in
Indo-European or Semitic to Sumerian.
Some linguists would say, for instance, that the contrast active - passive does not exist
in ergative languages. Karl Oberhuber, for example, in his examination of the Sumerian
passive, has said that: "Ein eigentliches 'Passivum
'
ist dem Sumerischen als einer
Ergativesprache von Haus aus Fremd" ( 1 982: 1 33). Earlier, Diakonoff said that ergative
languages "have no grammatical direct object, from which follows that ( 1 ) no Accusative
can exist; (2) no Passive and Active voice can exist" ( 1965: 1 8). However, Sumerian is
only split ergative; this means that such theoretical constraints may not apply equally to the
maru and to the gamt! (although not everyone agrees about the degree to which Sumerian
198 Lesson 1 7
is split-ergative).
The question also hinges on the nature of the Sumerian verbal root. Most
Sumerologists now say that the Sumerian rot is unmarked for voice or transitivity; that is,
du can be active or passive, tansitve or intansitive. These categories are not marked at all
in the rot; rather, they are determined by the syntax and semantcs of the entire sentence
within which they occur. The parallel has been made with English sentences of the type
"he is cooking". This sentence, on the surface, can either mean "the man is cooking-up",
because of the temperature, or "the man is coking some fo".
I any case, it seems that the two constuctions which are differentiated in English as
"intransitive" and "passive" are expressed by one constuction in Sumerian. (This is one
reason why some Sumerologists say that Sumerian has no passive.) Therefore, some
Sumerologists use the compound term "intansitive-passive" (or "passive-intransitive") to
refer to both constuctions. Other Sumerologists use the terms "passive" or "intansitive",
based on how the corresponding constuction in English (or German) comes out.
The difference between transitive and intransitive verbs can be illustrated using du, "to
build", and gin, "to go". In the bamty, these are:
( 1) The king went.
luga1.0 i. gin.0
(2) The king built the house.
lugal.e e.0 mu.n.du.0
In the intansitive sentence, the subject of the intransitive verb (the patient) is marked by .0.
This is cross-referenced by the .0 at the end of the verb. To express "the house was built",
which in English would be called a passive, Sumerian uses a construction identical with
sentence ( 1 ):
(3) The house was built.
e.0 Ldu.0
The .0 case-marker of the subject of the passive verb is cross-referenced by the .0 at the
end of the verb.
Sentence (3) is essentially the construction seen in line 17, although with the negative:
(4) gipar.bi. 0 nu.(L)du.0
nu is the same negative marker seen in Text 1 6. There, it occurred before a nominal
form (an active participle). Here it occurs before a verbal form. As stated in Lesson 1 6, nu
fts into the category of modal-prefx. As such, it is regulary followed by one of the
conjugation-prefxes. Here the conjugaton-prefx (.i ) has assimilated into the lul of nu.
The enclitic copula has occured several times, e.g., in the PN
d
Ba--nin-am, "Baba
is queen". In addition to being used in such equational sentences, am can also be used to
express circumstantial clauses. These can usually be tanslated into English as: "it being
that", "it being the case that", or by a participial phrase in -ing. When am is used in such a
constuction, it must follow a nominalized sentence. Hence the verb form is to be
understood as: [nu.(L)du.0] .a.am, meaning "it being the case that its giparu had not (yet)
been built".
To sum up, am is the enclitic copula, used here to express a circumstantial clause. .a
nominalizes the preceding sentence. The underlying sentence which has been nominalized
Lesson 1 7 199
is: gipar.bi.0 nu.(i).du.0.
18. The syntax of lines 17 and 1 8 is the same: am, following a nominalized sentence,
foning a circumstantial clause. In line 1 8, tH (intransitive in English) is construed just like
the verb du (passive in English) in line 1 7:
[gipar. bi.0 nu. (L)du.0].a.am
[en.0
nu.(i ).n(i).ti1.0] .a.am
The writing is slightly different: . . . -du-am vs . . . . -tll-la-am. This is presumably be
cause du ends in a vowel, but tU in a consonant.
The verb fon in line 1 8 differs in the presence of .n immediately before the verbal
root. This .n is somewhat diffcult to explain. It cannot be the personal-affx .n, because tU
"to live" is intansitive here: "no en-priestess had yet taken up residence" in the giparu.
More likely, this .n is a reduced fon of the dimensional-prefx which cross-references
the locative case. The usual form of this dimensional-prefx is ni. However, sporadic
instances of 1nl instead of Inil are attested; a possible case also occurs in Text 22. It has
not yet been possible to determine any phonological or morphological rules governing the
distribution of Inil and 1nl. It has been speculated that it has something to do with word
stess.
Just as the comitative dimensional-prefx da often occurs in a verbal prefx chain
without any corresponding comitative noun phrase for it to cross-reference, the locative
dimensional-prefx ni can also occur in verbal forms, without any corresponding locative
noun phrase. According to the studies of Gragg, such a use of ni often conveys a vaguely
adverbial force, and it can be translated simply as "there". It can refer loosely back to some
noun mentioned earlier, even if that noun is not in a locative case. In line 1 8, the reference
is to the giparu, the residence of the en-priestess, even though gipar.bi is not in a locatve
phrase (and, in fact, is more closely connected syntactically with the verb fon of line 1 7
than with the verb fon of line 1 8).
20. The .e is the ergative case-marker. In line 1 4, an ergative case-marker .e already
appeared. Since in general in Sumerian, case-markers appear at the end of a nominal
phrase, no matter how long the phrase may be, the presence of the ergative case-marker in
line 20 makes the presence of the ergative case-marker in line 1 4 "un grammatical".
However, it is very easy to see how this situation arose: lines 4- 14 fon a logical unit,
consisting of nothing but epithets of Amar-Sin. Lines 1 5- 1 8 are circumstantial clauses,
setting the background for Amar-Sin's activities. Lines 19-20 are almost a parenthetical
addition, a sudden shift of topic back to the agent of the sentence, repeating Amar-Sin's
name. Once these lines were "inserted", it was only natural (and perhaps necessary?) to
again add the ergative case-marker .e. In fact, some copies of this inscription also have an
ergative case-marker .e at the end of line 9. The scribe may have unconsciously been
infuenced by the fact that lines 1 - 1 4 fonn a complete unit by themselves. In several royal
inscriptions, the phrase lugal-an-ub-da-limmu
2
-ba-ke
4 marks the end of an agentive nomi
nal phrase. In fact, if lines 15-20 were omitted, the remaining lines 1 - 14, 21 -22 would have
foned a complete text by themselves.
21 . This is the direct object (patient) of the verb in line 22, "his (Nanna's) pure giparu".
22. This is the main verb governing lines 1 -21 . The essence of these frst 22 lines is:
200 Lesson 1 7
( 1 ) For Nanna benefactive
(4) Amar-Sin agent
( 1 5) -nothing having been built- circumstance
( 1 9) Amar-Sin agent
(21 ) his giparu patient
(22) built. verb
24. As discussed above, the verbal root in Sumerian is not marked for tansitivity; that is,
it can be used either tansitively or intansitvely, with no morphological change of the root
itself. For example, g (Text 6) can mean "to stand", or "to plant (a garden)". Thus, kur
9
can either mean "to enter", or, as in this paticular sentence, "to cause someone to enter", "to
bring in". The sense here is of installing the en-priestess in the giparu.
na is the dimensional-prefx which cross-references the dative, although no dative
phrase occurs. It is probably a kind of ethical dative or benefactive, loosely referring back
to Nanna, who is mentioned in lines 1 and 20.
ni is the full for of the dimensional-prefx cross-referencing the locative. Just as in
line 1 8, there is no locative noun phrase for it to cross-reference. It can be translated as
"there", referring loosely back to "his pure giparu" of line 2l .
I a sequence of two (or more) dimensional-prefxes, the dative dimensional-prefx
always comes frst.
25. It is writings such as these which show that a personal name of the type "Amar-Sin" is
indeed a genitive phrase. The writing presumably stands for amar. Zuen.ak.e. As discussed
in Lesson 1 , there is evidence that in certain (undetermined) conditions the .ak of personal
names was lost, and so the name was transcribed as "Amarzuen". Jacobsen, on the other
hand, even in his historical and literary publicatons, refers to this ruler as "Amarsuenak".
26. The four most common conjugation-prefxes in Sumerian are mu, L ba, and bi ; ex
amples of all of them have occurred. Besides these four, there are a certain number of
others, all with a Im/. The two most common are: im-ma and im-mi, with reduplicated
Im/. Others are written with one Im/: l-mi and l-ma. Others occur with different initial or
fnal vowels: am-ma.
The relatonship among these forms is unclear. It is not sure if these variations are
(mostly) orthographic (im-mi - i-mi), mostly phonological (im-ma - am-ma), or correspond
to a difference in meaning. These forms will be discussed further below.
In addition to these bi-syllabic forms, a form written im also occurs; this i s what is
written in line 26. Some Sumerologists believe that im is a reduced form of im-ma or im
mi (although the conditions govering such reduction are unknown). However, most
Sumerlogists think that a form such as im-da derives phonologically from libdal Lb.da;
this is the analysis followed here. In Text 17, for example, there occurs: im-da-ab-sud4 -
re
6
. Under this assumption, this would derive frm: Lb.da.b.sud.e.0. .i is the con
jugation-prefx, and .b is the optional prnominal-prefx which appears before the
comitative dimensional-prefx; here the reference is loosely ud, which is inanimate. The
phonological change may have been along the lines of I ibdal ) I iddal ) I imda/.
This explanation of im is plausible, because it is indeed normally followed by da or ta.
Lesson 1 7 201
If correct, then it should be understood as basically the conjugation-prefx i. There are
cases, however, where im is not followed by da or ta, and where some other explanation
may be necessary.
da is the comitative dimensional-prefx. It often ocurs with no corresponding comi
tative nominal phrase. Here it loosely refers back to ud, even though the latter is actually
cross-referenced as a patient.
Since the verb form is man, the .b before the verbal root cross-references the direct
object ud. The . 0 after the man-suffx cross-references the subject Amarsin.ak.e:
Amarsin.ak.e ud.0 Lb.da.b.sud.e.0
As was just said above, the ud is directly cross-referenced by the .b before the verbal
root, but it is also loosely cross-referenced by the comitative dimensional-prefx da, with its
(assumed) inanimate pronominal-prefx .b.
sud4 is a member of the afxation class, forming its man by addition of the man-suf
fx .e.
To summarze the verb form:
i . b . da
( 1 ) (2) (3)
b . sud . e . 0
(4) (5) (6) (7)
( 1 ) conjugation-prefx
(2) optional pronominal-prefx
(3) comitative dimensional-prefx
(4) personal-affx cross-referencing patent
(5) verbal root
(6) maru-sufx
(7) personal-afx cross-referencing agent
Discussion: stucture
The bare-bones stucture of this text is:
1 -3 For Nanna benefactive
4- 14 Amar-Sin agent
15 i n Karzida place
1 6 from of old time
17 no giparu having been built circumstance
1 8 no en having lived circumstance
1 9-20 Amar-Sin agent
21 his giparu patient
22 built. verb
23 His en patient
24 he made enter. verb
25 Amar-Sin agent
26 its days patient
26 will prolong. verb
202
27
28
For his life
he dedicated.
Lesson 1 7
purpose
verb
- Passive
Ergativity is defned by the way the relationships among the primary paticipants in a
sentence are marked. In the bamt!, which functions on an ergative basis, the subject of a
tansitive verb is marked by .e. The subject of an intansitve or passive verb, and the direct
object of an active verb, are marked by .0. (The man of intansitive-passive verbs will be
discussed later.)
I Sumerian,there is no difference in constuction between an intransitive sentence and
a passive sentence in the bamt!:
luga1.0 Lgin.0
e.0 Ldu.0
Perhaps the most cautious statement about the passive in Sumerian is that of Jacobsen:
The exteral criteria detenining whether a Sumerian fon is active or
passive in meaning are as yet far frm clear and the whole question whether
in actual fact this distncton may be considered genane to the Sumerian
verb is yet to be decided ( 1956:49*).
Similarly, Viktor Christian, many of whose ideas about S umerian grammar are outside
the main-stream, says:
In Sumerian, we do not fnd the categories "tansitive" and "intransitive", or
"active" and "passive". The fact that we are often forced to translate as active
or passive, tansitve or intansitve, only results from the inadequacies of our
own language to corectly reproduce S umerian (1961 : 1 3).
That is, the contasts active - passive, and tansitive - intansitive, may not be the most
ftting way to describe Sumerian. Some Sumerlogists have stated that an analysis in tens
of action - state would more fttingly describe Sumerian. In practice, however, people
sometimes tend to be less than rigid in their use of such linguistic tens.
Traditionally, Sumerian grammars present two paradigms: one for the active; and one
for the intansitive-passive. There is no difference in the stucture of the root; rather, the
differences lie in the function of the personal-affxes, and in the way that the case-endings
of the primary paticipants are cross-referenced. It may be useful to summarize the bamm
fons in the singular.
frst person
second
third animate
inanimate
-Conjugation-prefxes
active
mu.0. sar
mu.e. sar
mu.n. sar
mu.b. sar
intansitive-passive
mu. sar.en
mu.sar.en
mu. sar.0
mu.sar.0
The relationship of the conjugation-prefxes im-mi, im-ma, im, and other rarer fons is
unsure. It is usually assumed that these are not unanalyzable, unitary morphemes. Rather,
Lesson 1 7 203
many Sumerologists assume that the two conjugation-prefxes im-mi and im-ma somehow
derive from bi and ba (although it is not always made clear if this derivation is to be
understoo i synchronic or diachronic terms). Falkenstein, for example, who has a
different understanding of the conjugation-prefxes than that presented in this book, derives
im-mi from *i-bi. Others, however, have argued for the existence of a morpheme fmf in all
these forms. It would rughly corespond in function to the Akkadian ventive (which is
normally used with verbs of motion, marking movement towards the speaker).
Such historical developments are unprven, and perhaps unprvable; Falkenstein, for
example, had to posit a number of apparently unmotivated phonetic changes to get his
forms to work. In general, some Sumerlogists (such as Falkenstein) are inclined to see
variation in Sumerian as due to phonetic reasons, even if the rules governing the phonetic
changes cannot be determined; others (such as Jacobsen) are inclined to believe that
Sumerian grammar is more complex than usually believed, and that the variaton we see is
due to our ignorance of the morhology, not to unexplained phonetic accidents. In any
case, in synchronic terms im-ma and im-mi (and i patter the same as the other conju
gation-prefxes; that is, they all occupy the same position in the verbal prefx chain, and
their presence is mutually exclusive. More work remains to be done on the synchronic
distribution ofthese conjugation-prefxes. Krecher ( 1985) is the latest attempt to isolate the
morphology and the semantcs of the various conjugation-prefxes in fmf. He posits a
rather wide range of functions, which need to be more fully investigated.
-Case-markers
The presence of the two ergative case-markers (lines 14 and 20) has parallels in other
agglutinative languages. If a constuction starts to become very long or convoluted, the
speaker (or writer) will occasionally "get lost" in the construction, and may occasionally
back-tack, changing the topic, and will have to repeat a previous case-marker.
- History
The giparu at Ur was the "ofcial" dwelling-place of the en-priestess (who is some
times refered to by the Akkadianized term "en tu-priestess"). It was a large structre,
composed of many rooms. The frst such stucture at Ur may go back to Early Dynastic
times; it was built and rebuilt right through the Neo-Babylonian period.
The en-priestess was always of royal blood. Perhaps the most famous was
Enheduanna, the daughter of Sargon of Akkad. She is well-known as the author of two
well-preserved poems, written in good Sumerian. In the Neo-Babylonian period, Nabo
nidus installed his daughter in the positon.
The en-priestess "represented" the goddess Ningal, in some way. In particular, she
represented the goddess Ningal, while the reigning monarch represented the god Nanna
(the husband of Ningal) in some kind of "divine marriage ceremony". This ritual may have
originated as an end-of-harvest-time festival. It has ofen been discussed among Sumero
logists, and there has been much disagreement about what the sacred marriage was meant to
represent.
Penelope Weadok has summarized the functions of the giparu:
20 Lesson 17
Three separate units emerge from the Ur III--Isin-Larsa giparu building: the
Ningal temple which is the locale in which the entu-priestess, as the incar
nation of the goddess Ningal, carried out her most important function as a
participant in the rite of the sacred marriage; the giparu proper which was the
offcial dwelling of the entu-priestess, with its annexe, the cemetery for the
former entus; and the sanctuary in which the entu prayed for the life of the
king, her father or brother, in the hope that the gods would bestow prosperity
upon the land through the king, their human regent. ( 1 975: 1 24).
It has been argued that the "institution" of the giparu existed in other cities, for ex
ample, in Uruk. This is probably tue, but the evidence is not conclusive. However, Tex1
17 evidently refers to a giparu in Karzida, not to the giparu in Ur. This implies a Nanna
temple in Gaesh. Nothing is known of this temple or this giparu.
The orginal meaning of the term -par is unknown; it may have been "storehouse
'
of some kind.
L0880u 1 8
This is a dor soket of Shu-Sin, the brother and successor of Ama-Sin. He ruled
from 2037 to 2029 BC.
Sign-list and vocabular
"'
An-nu-ni-tum Annunitum (DN, masc)
'-
Su-
d
Zuen Shu-Sin (PN)
dam wife, consort
Notes
An-nu-ni-tum This may originally have been an epithet of Inanna. In Sargonic texts, the
divine name
d
lnanna-An-nu-ni-tum is ocasionally attested. But after the Old Akkadian
period, this compound term does not occur, only the individual term An-nu-ni-tum. Ac
cording to Jack Roberts, this patter of attestation "suggests that the epithet split off and
became an independent deity" ( 1972: 147).
Gelb has pointed out "the temendous number of compound divine names in the Ur III
period" ( 1987: 125). The name
d
lnanna-An-nu-ni-tum belongs to the class of names cha
racterized by Gelb as "DN plus description".
The etymology is unknown. It is possible that the -itum ending is an Akkadian femi
nine gentilic: II-t-um/. However, the meaning of "anun" or "annun" is unknown; it is not
sure ifit is Akkadian or Sumerian.
S
u-
d
Zuen The name is Akkadian, meaning "the one of Sin", or "the one belonging to
Sin". It was formery read as "Gimil-Sin".
205
206
Lesson 18
Text 1 8
1. The name of the goddess is clearly An-nu-ni-tum.
Lesson 18 207
Notes
Many objects containing inscriptions have been damaged, either in ancient times or in
modern times. This means that part of the inscription may be completely broken away, as
in the bottom right-hand corer of Text 17, or part of it may b effaced, as is the top left
hand corner, and as in case 8.
It is important that tansliterations refect such damage, especially when the text is
completely broken. Unfortunately, there is no one system in use by Sumerologists (or
Assyriologists) to indicate such breaks. The most common system is to use brackets.
These are used to include signs completely broken away. Thus, line 1 1 is best tans
literated: -- [ni]. Brackets can also be used to indicate partially broken signs. For
example, line 8 can be transliterated: lugal-[kala]g-g, and the last line as: mu-n[-du] .
Brackets are imprecise, however, in such cases as the partially effaced An-sign in line
. To resolve this problem, some scholars (not all) use half-brackets to indicate partially
broken signs. Thus, the frst sign of line 1 can be transliterated rAn' .
Instead of half-brackets, some scholars use dots under vowels, to indicate that the sign
is partially broken or effaced. Thus, this An-sign might be transliterated as An. Some
scholars use different combinations of full brackets, half-brackets, and dots, in sometimes
rather idiosyncratic combinations.
The decision about whether or not to use brackets (or half-brackets) is not always
clear-cut. For example, what about the tur-sign in line 1 , or the frst part of the Urims-sign
in line 9? In practice, such damage to the text is often ignored, if the context and the
remaining traces of the sign make the sign unambiguous.
Similarly, it is diffcult to decide how breaks should be refected in translation. Some
Sumerologists use brackets in translation, refecting the breaks in the text. However, since
English and Sumerian are of such different grammatical stuctures (particularly in word
order), this procedure can be cumbersome and tiresome. Many scholars, therefore, omit
brackets i

translation, especially when writing for a professional audience, because such
an audience will be able to follow or control the tansliteration or autograph.
Occasionally, scholars may be suspicious of a published tansliteration, or even an
autograph, of a cuneiform text. O, they may wish further information about a partially
effaced sign. In such cases, they may collate the text, that is, physically examine the
cuneiform document. If the text is not easily accessible, they may ask another scholar to
perform such a collation.
The system used here is the simplest. Full brackets are used only to indicate
signifcant breaks. They are omitted fom transcription and fom tanslation.
Transliteration Transcription Translation
1 : An-nu-ni-tum Annunitum For Annunitum,
2: dam--ni-i dam.anLr his wife -
3:
d
Su-
d
Zuen Susin Shu-Sin,
4: ki-
2
-
d
En-lil-1 kLaga.a.Enlil.a(k) the beloved of Enlil,
2
0
8
Lesson 18
5: lugal-
d
En-lil-le lugal Enlil.e
6: ki-
2
-Sag4-g-na kLaga.0.Sag.anLa
7: in-pad Ln.pad
8: luga1-[kala]g-g lugal.kalaga
9: lugil-Urims
ki
-ma lugal.Urim.a(k)
10: lugal-an-ub-da-limmu
Z
lugal.anub.da. (k) limmu.bLak.e
oa:e4
1 1 : --[n
12: mU-!@-.l
Commenta
e.anL0
mu.na.(n.)du.0
the king whom Enlil
selectea in his
loving heart,
the mighty king,
the king of Ur,
the king of the four quarters-
her temple
he built.
1 . Lines 5-7 are an epithet of Shu-Sin, which occurs in virtually all of his inscriptions. It
was not used by his predecessors, nor by any subsequent ruler.
Although this epithet is very common, and its basic meaning is fairly transparent, the
syntax underlying it is not clear.
In particular, the grammatical relationship ofki-g to Sag4 -g-na is not sure. One in
terpretation of these lines is: "the king whom Enlil has elected as the beloved of his heart",
that is, kLaga.a Sag.anLa(k). However, Sumerian would probably use -Se to express the
complement introduced in English by "as"; cf. Text 19, where -Se is used in a (roughly)
parallel context.
Another interpretation is to see ki-
2
as an active participle modifying Sag4
'
with the
entire noun phrse being in the locative: "in his loving heart", that is, [kLaga.0 Sag.ani].a.
The problem with this interpretation is that modifers of nouns almost always follow their
nouns, not precede them; one would not expect kLaga.0 to precede Sag.anL However,
there are exceptions to this rule, and there are even occurrences where simple adjectives
precede their nouns (although these occurrences are mostly of a formulaic nature). There is
probably some stylistic emphasis present.
Lines 5-7 consist of the noun lugal and a relative clause. Since this is a relative clause,
one would expect to fnd a sentence nominalized in .a, as was the case in all other relative
clauses with a verb. However, this common epithet is always written in-pad, with no
nominalizing .a. This is diffcult to explain. It is not simply a problem of orthography. A
follower of the Falkenstein school might wonder about the possibility of a reading /pada/
for the pad-sign. However, no such value is recognized by the standard sign-lists, and in
any case, one would expect it to be written in-pad-da (as in Amar-Sin
'
s epithet, Enlil.e
Nibru.a mu.pad.a, always written mu-pad-da).
Somewhat similar instances of relative clauses without an expected nominalizer occur
elsewhere in S\merian, sometimes of a formulaic nature. They need further investigation.
7. i is the conjugation-prefx. n. is the personal-affx cross-referencing the U..t!-agent,
Enlil.e.
Lesson 18
Discussion: structure
The stucture of this text is:
[Annunitum, dam.ani].r benefactive
[Susin, ki.aga.a.Enlil.a(k), agent
lugal Enlil.e ki.aga.0.ag.anLa tn.pad,
lugal.kalaga, lugal. Urim.a(k),
lugal.anub.da.(k) limmu. bLak] .e
[e.ani] . 0
mu.n.(n.)du.0
patient
verb
209
It thus follows the patter of most royal inscriptions. This basic patter, however, is
somewhat difcult to recognize, because of the length and the complexity of the
appositonal phrases in lines 4 through 1 0.
-Relative clauses
The interpretation of lines 5-7 given above follows Jacobsen, who tanslates this
formulaic phrase as "the king whom Enlil envisaged in his loving heart". He considers it to
be a clause nominalized "in zero"; however, he cannot fd many close parallels. These
lines illustrate a problem encountered several times before: a construction ocuring fre
quently, its meaning relatively tansparent, but its syntax dubious. There are still severl
problems in the understanding of Sumerian relative clauses.
-Adjectives
There are a few cases in Sumerian where adjectves (or other modifers) precede their
head noun, instead of following it. For example, the adjective kug, meaning "pure", quite
frequently precedes the names of gods and goddesses: kug-
d
Inanna, "pure Inanna". It ap
pears to be the only adjective to be used so regularly in this position.
Other languages whose order is basically noun-adjective, such as French, also permit a
certain number of cases of adjective-noun constructions. These cases are usually limited to
a fxed number of adjectives or expressions. In general, it seems that languages of the
noun-adjective type permit more exceptions than do languages ofthe adjective-noun type.
- History
The circumstances of Amar-Sin
'
s death are unclear. A late omen text says that he died
of an infection caused by a foot bite of some kind. Shu-Sin was his son and successor
(although it has been said that he was his brother, not his son). It was during the latter
'
s
reign that trouble began to be felt in the empire: the Sumerians began to feel the pressure of
the Amori tes.
Curiously, a number of "love poems" (to use Jacobsen
'
s term) have been preserved,
most of which are directed to the fourth king of the dynasty, Shu-Suen. One
guesses that this king, or perhaps more likely his queen, had in his entourage
a woman poet who enjoyed singing about love and lovemaking, and whose
210 Lesson 18
works, since they were cast in the fon of praise for the king
'
s beauty and
virle prowess, were favorably received and carefully preserved in writing
( 1987: 85).
Text 1 8a
supplementary
This is a votive bead of carnelian, found at Susa. It has been speculated that this bead
was caried off to Susa as booty by the Elamites when they sacked Ur in 2004 BC, putting
an end to the Ur III Dynasty.
- |


-
L0880u 1 9
This door socket of Shu-Sin records one of the more signifcant events of his reign.
Sign-list ad vocabular
Sara
2
Shara (DN, masc)
Nin-lil
--
Ninlil (DN, fer)
the Amorites; the West
}- Mu-
n-iq Ti-id-ni-im Muriq-Tidnim (GN)
! ' s e-ag4 -g-r-da Eshagepada (T)
_ nir-gaI prince
_ ad-da father
1-
gib (kind of priest)

gudug (gudu4) (kind of priest)


!
u hand
_ sipad (sipa) shepherd
ne forces, troops
_ ma-da land

dadag to be clean, pure

gal great
Notes
Sa God of the city of Umma; son of Inanna. Not much is known of him; he did not
rank very high in the Sumerian pantheon.
Nin-l1l According to Sumerian mythology, Ninlil was raped by Enlil, and later became his
wife (dar). Nanna was their child. She was especially worshipped in Nippur.
7J
212 Lesson 19
Mar-tu This term was used by the Sumerians in two ways. In an "ethnic" sense, it refers
to the (West-Semitic) Amorites, who dwelt in the Syrian desert west of Mesopotamia. It is
also used in a rather vague geographical sense, meaning "west" in general.
The etymology is unknown. Some now read it as Mar-du.
Mu-r-iq Ti-id-ni-im The name is Akkadian, Murq Tidnim, meaning "that which keeps
Tidnum away". murq is the D-stem actve participle in the construct state, from req!, "to
be far away". Tidnum (or "Didnum") is probably the name of a particular Amorite tibe,
although here it refers to the Amorites in general. In late lexical texts, "Tidnum" is equated
with the Akkadian word used for the Amorites: r-id-nu !-mur-ru-y. The original form
and etymology of the name are unsure. In the various Semitic languages, several variant
spellings of the name Tidnum occur: Ti-da-nu-um and Di-ta-nu-um in Akkadian, Ddn and
Ttn in Ugaritic, etc. It is not always sure if such terms refer to the same people.
E
-ag
4
-ge-pAd-da The principal temple of Shara in Umma. The name means "the temple
chosen in (his) heart", that is, e.ag.e.pad.a. The . e is the mark of the locative-terminative
case.
nir-g8 It is diffcult to determine the precise meaning of such a word; it is conventionally
tanslated into English as "prince" or "hero". The Akkadian equivalent is etellu, glossed by
the CAD as: "prince, lord".
In origin, this is probably an active participle with an incorporated object. gal normally
means "to be", but can also mean "to have". The combination nir-gal would mean "the one
who has nir". Unfortunately, it is not known what nir means.
ad-da Sumerian has three words for father: !-!, ab-ba, and ad-da. Lambert ( 1 957) stu
died the distribution of the three terms on (primarily) geographical and class lines, but could
not come to any defnite conclusions about their usage.
iib This is a very old loan-word into Sumerian from Akkadian waSipu. This Akkadian
term is usually tanslated "exorcist". iSib was then loaned back into Akkadian as iippu,
glossed by the CAD as "purifcation priest". iib itself is often translated as "incantation
priest".
It is never easy to determine the exact function of any priestly offce. Iohannes Renger
( 1 967f has studied the Old Babylonian priesthood, and exhaustively discussed the
Akkadian words for the different kinds of priests.
gudug The equivalent Akkadian priest is the paSgu, from the root paSau, meaning "to
anoint". The gudug-priest seems to be especially involved in certain kinds of ritual
activities, which it is not yet possible to defne exactly. He was of a lower rank then the
iib-priest (at last in the Old Babylonian period); temples could have several gudug-priests,
but apparently no more than one iSib-priest.
This word is often tansliterated as guda4
'
with the second vowel as lal, not as lu/.
sipad The cuneiform sign used to represent this word is actually a combination of two
Lesson 19
2
13
signs: the Q-sign ( *) and the udu-sign _, The Q-sign was originally a picture of a
staff or scepte of some kind; in this meaning, it is read as gidri. It was also used to
represent the word for "overseer", Le. , a man holding a staff of authority; in this meaning, it
is read as !gula. The udu-sign means "sheep". Thus, the sipad-sign graphically represents
approximately "the overseer in charge of sheep".
It is reasonably sure that this word has a Id/-Auslaut. However, it is much more
common to fnd it transliterated as sipa.
ma-da This is usually assumed to be a very early loan into Sumerian from Akkadian matu
(although the latter is itself of dubious etymology; the only other Semitic language it occurs
in is Aramaic). The three terms kalam, ma-da, and kur are often tanslated into English as
"land", but they are not synonymous. Limet (1978) has studied their distribution, espe
cially in documents from the Ur III period. kalam is used exclusively to refer to Sumer.
kur originally meant "mountain". It then came to mean "foreign land"; it is never used to
refer to the land of Sumer. ma-da is more problematical; the sense is approximately
"territory". It is used mostly for foreign lands, but in certain uses it can refer to Sumer.
Jacobsen believes that in contexts such as Text 19 ma-da means "steppe". The oppositon
between kalam and kur has also been studied by Steiner (1978) on a number of levels:
historical, legal, etc.
Because the term kalam is restricted to the land of Sumer, Poebel has speculated that
the word Ki-en-gi was a dialectal form of kalam. ka-na-.M is used as the Emesal equivalent
of both Ki-en-gi and kalam. However, in earlier texts kalam may have had a more general
meaning. Raphael Kutscher says that "Although the literal meaning of kalam is 'country
'
(matum), it narrowed its scope to 'The Country
'
par excellence, namely, Sumer, and even
tually, to 'the nation', Le., the Sumerians" ( 1 975: 68).
dadag The reading is not certain; it is variously read as: da7 -dag, dag-dag, zalag-zalag,
zalzalag, and babbar
2
' The reading as I dadagl seems established by late lexical texts,
which give the syllabic writing da-da-fg as the equivalent of the Akkadian word for "pure",
ebbu.
The dadag-sign is formed by the writing of two signs; the situation is similar to
that of the siskur1 -sign. The word Idadagl is probably a reduction of something like
*/dagdag/, a reduplicated form. The reduplication may have had some kind of intensive
value in its origin. This explains why there are two signs. Originally, one sign stood for
Idag/; two signs stood for Idagdag/. Later, */dagdagl was reduced to Idadag/, but the
original two signs continued to be written. Similarly, it has been speculated that the word
Isiskurl was originally a reduplicated form, perhaps */sikur-sikur/; this explains why it
was written with two signs. One would then have to assume a development along the lines
of */sikursikur/ ) */sisikur/ ) Isiskur/, or something similar.
2
14
Lesson 19
0Xl 1 V
J

` t Rest mistake ofscribe.


Lesson 19
2
1
5
Notes
Line 23: As the editors point out, the scribe has drawn a ni-sign, instead of the expected
du-sign. He may have been unconsciously infuenced by the ni-sign directly above, in line
22.
1 :
2:
3:
4:
5:
6:
7:
8:
9:
1 0:
1 1 :
12:
1 3:
14:
15:
1 6:
17:
1 8:
19:
20:
21 :
22:
23:
24:
25:
26:
27:
Transliteration
d
Sara
-
2
nir-M-An-na
- --
dumu-ki-g
2
d
lnanna
ad-da-ni-i
d
Su-
d
Zuen
mb-An-na
gudug-u-dadag
d
En-HI
d
Nin-lil-ka
.! dig.-gal-gal--ne
lugal
d
En-lil-Ie
ki-g
2
ag4--na
in-pad
sipad-kalam-ma-e
lugal-kalag-g
I I U
ki
-
-g - nm
S
-ma
lugal-an-ub-da-limmuZba-ke4
ud bad-Mar-tu
Mu-n-iq
Ti-id-ni-im
mu-du-
.!ne-Mar-tu
ma-da-ne-
bi-in-g4-
E-ag4-g-pad-da
Transcription
Sara
nirgal.An. a(k)
dumu.kLaga.a
Inanna.(k)
adda.anLr
Susin
iib.An.a(k)
gudug.u.dadag
Enlil
Ninlil. (a)k.a(k)
u digir.gal.gal.ene. (k)
lugal Enlil.e
kLaga.0
ag.anLa
Ln.pad
sipad.kalam. a(k). e
lugal.kalaga
lugal.Urim.a(k)
lugal.anub.da. (k) limmu.bLak.e
ud bad.Martu.(k)
Muriq
Tidnim.0
mu.(n.)du.0.a.a
u ne. Martu. (k.)0
mada. anLe
bLn.gL0.a.a
Eagepada
216
28: -ki-

g-M-ni
29: nam-ti1-la-ni-e
30: mu-na-du
Translation
1 : For Shara,
2: the prince of An,
3: the beloved son of Inanna,
5: his father -
6: Shu-Sin,
7: the iib-priest of An,
Lesson 19
8: the gudug-priest with the pure hands
9: ofEnlil and Ninlil
1 1 : and of the great gods,
e.kLaga.a.ani.0
nam.til.anUe
mu.na.(n.)du.0
1 2: the king whom Enlil selected in his loving heart
1 6: to b shepherd of the land,
17: the mighty king,
1 8: the king ofUr,
19: the king of the four quarters -
23: when he built
20: the Martu-wall
21 : (whose name is) Muriq-Tidnim,
26: and when he drove back
24: the forces of the Martu
25: to their own land -
30: he built -
27: theEshagepada,
28: his beloved temple -
29: for the sake of his life.
Comentary
8. dadag is an adjectve from the root "to be pure".
The presence of the ka-sign in line 10 means that the construction is a double genitive:
"the gudug-priest of the pure hand(s) of Enlil and Ninlil, and of the great gods". It is
difcult to say exactly what this means. However, the "clean hands" of the paBu-priest
Lesson 19 217
are explicitly mentioned several times in Akkadian texts. The sense of the sentence is
probably: "the gudug-priest with the pure hands, in the service of Enlil and Ninlil, and in
the service of the great gods".
9- 1 1 . Enlil and Ninlil are conjoined, without any conjunction. These two deities are set
offfrom the more vague "great gods" by the conjunction y.
The two plural formations of nouns seen so far are the morheme .ene (for animate
nouns) , and reduplication (for inanimate nouns). Another common plural formation, used
when a noun is followed by an adjective, is to reduplicate the adjective: digi-gal-gal. This
type of foraton is most common with animate nouns, but a few examples occur with
inanimate nouns: inim-gal-gal, "the great words". In addition, it is possible to tack on the
plural marker .ene: digir. gal.gal.ene, which is the form in Text 19.
Plural formations such as digi-gal-gal and digi-gal-gal--ne are usually thought to be
superlatives of some type: "the most great gods". This is very hard to prove. It is
especially hard to determine if there is any difference between the reduplicated forms with
.ene and the reduplicated forms without it. Only a few adjectives seem to occur in such
plural formations; by far the most common is gal.
In Text 6, the plural "king of the gods" was written ilgal-dgir-re- with redup
lication of the fnal Irl of di. Line 1 1 of Text 1 9 uses a more morhemic writing; this is
the regular writing of this particular expression.
16. One of the functions of the terminative case in -e is to express the purose of certain
verbs. Here the sense is: "selected to be the shepherd of the land". Expressions of this
type are not infequent in the royal inscriptions. In them, the nominal phrase marked in -e
regularly follows the verbal form in-pad, instead of preceding it.
This deviation frm standard Sumerian syntax is presumably to give some degree of
emphasis to this last constituent. There are instances where adverbial phrases, and even
patients, are placed afer the verb phrase, instead of before it.
19. The .e marks the end of the ergative agent phrase, spanning lines 6 through 19.
20. Lines 20-26 form two when-clauses, governed by the ud of line 20:
20: when he built . . .
24: and when he drove back . . .
30: (then) he built . . . .
The two when-clauses are linked by the conjunction y in line 24. The syntax of these
clauses is the same as that seen in Text 1 2. ud is the relative marker; it is followed by two
clauses, each of which is nominalized in .a, and each of which is in the locative case. A
more literal tanslation would thus be: "on ( the locative .a) the day that ( the nornalizer
.a) . . . ":
ud [bad. Martu ... mu.(n.)du.0.a].a
u [ne.Martu ... bLn.gi.0.a].a
bad-Mar-tu is presumably a genitive phrase, "the wall of Martu". Since Mar-tu can
also mean "the West", this is sometimes tanslated as "the Western Wall". What follows is
the actual name of the wall. Instead of using a construction with the word for "name" (mu),
the name is expressed through an appositive. Such formulations are quite common.
24. !links the two temporal clauses, each dependent on the relative marker ud:
218
ud: 1) . . . mu-du-
2) . . . bi -in-g4-f
Lesson 19
25 . . ani is somewhat ambiguous. It could refer back (loosely) to Shu-Sin, or it could refer
back to the ne-Mar-tu of line 24, treating the term as a singular or a collective .
. e is the marker of the locative-ternatve case. As discussed in Lesson 9, this case
shares some of the characteristcs of the locative case marked in -, and some of the
characteristics of the terminative case marked in -e. This can lead to a certain amount of
ambiguity. For example, this particular line has been interpreted in two ways. One
interpretation is to understand .e here in the sense of -e. The meaning would then be: "he
drove the Amorites back to their own territory". However, it is also possible to understand
.e in the sense of -f; the meaning then is: "in his own territory, he drove out the Amorites".
As was also the case in Text 9, the locative-terminative is not cross-referenced by any
dimensional-prefx in the verbal chain.
Line 25 has been transliterated here as ma-da-ne-. This assumes that the lit of lanil
has contracted into the lel of the locative-terminative case-marker, producing a pronun
ciation something like Imadane/. Other scholars have transliterated the line as ma-da-ni-.
Some Sumerologists do this, because they do not believe that such a contraction took place.
Others do it, because they are consciously being morphemic in their tansliteration. This
problem is further discussed in Lesson 21 .
26. In the verb i n line 23, the personal-affx . n for the bamt!-agent is not expressed in the
writing. However, in line 26 it is so expressed:
23: mu-du-f
26: bi -in-g4-i.
It is also not expressed in the verb in line 30:
30: mu-na-du.
Frm this one text, a facile generalization would be: The personal-affx . n is not used
when the conjugation-prefx mu is present, but it is used when the conjugation-prefx bi is
present. However, in other texts the writing mu-na-an-du occurred. It is such varying for
mulations which makes it very diffcult to determine the rules governing the presence or ab
sence of the personal-affx.
Discussion: structure
Although this inscription is thirty lines long, it consists of only one sentence. The one
fnite verb form is in line 30. It may help to summarize the structure of the text:
1 For Shara, benefactive
6 Shu-Sin, agent
20 - when he built . . . - circumstance
30 he built verb
27 theEshagepada patient
29 for the sake of his life. purpose
Lesson 19 219
It is not uncommon to fnd a verb with a bi conjugation-prefx co-occurring with a
nominal phrase in the locative-terminative case. In Text 19, line 26 uses the conjugation
prefx h and line 25 has a nominal phrase in the locative-terminative. Earlier it was
mentioned that it is also not uncommon to fnd a verb with a ba conjugation-prefx co
occurring with a nominal phrase in the locative case. The signifcance of these facts still
remains to be integrated into a coherent theory of Sumerian grammar.
- Loan words
Both iib and ma-da are early loan words fom Akkadian. One, however, ends in lal,
and one doesn
'
t. Gelb has speculated that such very early loans refect a stage of Akkadian
when the case-system was not as fully developed as it was during the Classical periods.
However, it has also been suggested that the lal is the Sumerian nominalizer.
- Chronology
As the name of this wall implies, its function was to keep away the nomadic Amorites.
Individual Amorites, or small groups, had been entering into southern Mesopotamia from
the North-west for many years, but during the reign of Shu-Sin they began to enter in
force.
The building of this wall is also mentioned in a "year-date" of Shu-Sin. Until the
Seleucid period, there was no chronological system in Mesopotamia based on a fxed date.
Dating systems varied from place to place and from time to time. Beginning at least as
early as the Old Akkadian period, individual years in a king
'
s rule were given their own
names. For example, the frst year of the rule of Shu-Sin was named: "(the year when)
Shu-Sin became king"; his third year was named: "(the year when) Simanum was
destroyed". That is, the name given to the year referred to some important event in the rule
of the king. These year-names were gathered into lists. Without these ordered lists, it
would be impossible to know which particular chronological year a given year-name
referred to.
Year 4 of Shu-Sin
's rule is named
d
Su-
d
Zuen ilgal Urims
ki
-ma-ke4 bad-Mar-tu Mu
ri-iq-Ti-id-ni-im mu-du, that is, "(the year when) Shu-Sin, king ofUr, built the wall against
the Amorites, (named) Muriq-Tidnim". The wording of this year-name is similar to the
wording in lines 20 through 23 of Text 19. Douglas Frayne has said: "It can be
demonstated that temporal clauses in royal inscriptions of the Ur III through Old
Babylonian periods often allude to year formulae of the king" ( 1 983:745); this is apparently
such an instance.
In a similar vein, Hallo has pointed out:
The correlation between neo-Sumerian regnal years on the one hand and
royal hymns on the other is a high one both in terms of numbers and in terms
of content. . . . Is it too daring to suggest that each date formula was formally
intoduced together with a new hymn? ( 1 966: 1 39, and n. 82).
The year-formula just cited is the "long" form of the year-name; there is also a "short"
form: bad-Mar-tu ba-du, "(the year when) the wall of Martu was built". This latter is a
220 Lesson 19
passive constuction, bad. Martu. (k). 0 ba.du.0.
Year-dates often occur in both a long fon and a short fon. The long fon has an
agent marked in .e, and a verb with conjugation-prefx mu. The short fon is without
agent, and has a verb with conjugation-prefx ba. For example, there are several year-dates
of the type: mu PN.e GN mu-bul, "the year when PN destoyed GN", and mu GN ba-bul,
"the year when GN was destroyed". In each case, the verbal sentence is in apposition with
the noun mu "year".
The conjugation-prefx ba fequently occurs with verb fons in passive sentences,
such as the short fon of year-dates. In the Old Babylonian Grammatical Texts, Sumerian
verbal fons with the conjugation-prefx ba are usually translated by the Akkadian sters
with infxed It I , that is, sters with separative, refexive, and passive meanings. However,
Malcolm Horsnell ( 1 977) has questioned the standard interpretation of the short fon of
year-dates as passives, arguing that they should be translated as agentless active sentences.
It has also been speculated that there is not just one conjugation-prefx ba, but rather
there are two; that is, they are homonyms. One is seen chiefly in passive sentences, the
other in less defnable contexts. Needless to say, it is very hard to prove such an assertion.
-Literary parallels
Other references to this wall have been preserved. One is in a letter by Sham-bani,
the ofcial in charge of its construction; he writes to Shu-Sin complaining of his troubles.
Shu-Sin
'
s own reply to Sharrum-bani, in which he berates him for neglecting his duties, is
also preserved.
At frst blush, it might strike one as rather astounding that such letters just happen to
be preserved. But the reason is because these (and other) letters came to be considered
literary texts, and were used for scribal practice. According to the defnition of
Michalowski, "royal literary letters are thus simply letters to and from kings which were
recopied in the scribal academy as part of the instructon in learning the S umerian language"
( 1 980b:52). Similarly, K.R. Veenhof says:
A number of letters of kings of the Third Dynasty of Ur and some of their
successors were studied and copied in the Old Babylonian schools and
became part of the standard curiculum of that period ( 1 986:6).
Most of these royal literary letters are products of the Ur III period. They are only
known frm Old Babylonian copies; fragments of at least four copies of the letter frm
Sharrum-bani to Shu-Sin are preserved.
- History
It is possible that Shu-Sin's activities did not consist of the building of a "wall" de
novo; it may have been the rebuilding of defensive fortifcations started by Shulgi. Accor
ding to one letter (unfortunately, somewhat fragmentary), this wall was designed to be "26
double-hours
'
: long, that is, about 170 miles! Attempts have been made to relate this wall
to defensive lines mentioned in other texts, and even in Classical sources, and to detenine
the location of the wall, but such attempts have lacked conviction.
The wall was ultimately unsuccessful. c.J. Gadd says:
Lesson 19
As for his great wall, it proved even more ineffectual than such barriers have
always been in the end. No more is heard of this vast and vain work, even
if, as seems likely, it furnished a line or a foundation for similar works in
later ages. Babylonia has no natural defences, and they were not to be pro
vided by an artifcial rampart so long that it could have hardly been
effectually garrisoned ( 1971 : 61 1 ).
221
It seems curious that this wall was given an Akkadian name, instead of a Sumerian
one. This may attest to the growing importance of Akkadian as the spoken language during
the Ur III period. Possibly it was chosen because its meaning would have been more-or
less understandable to Amorite speakers.
- History
The ninth year of Shu-Sin commemorates the building of the temple mentioned in this
inscription: mu _
d
Sara
T
Umma
ki
-ka ba-du, "the year when the temple of Shara of Umma
was built".
Such year-dates, as laconic as they are, actually comprise one of the principal sources
of information about the history of the Ur III period.
Text 1 9a
supplementary
This text was inscribed on a votive amulet of agate.
222 Lesson 19
Sign-list ad vocabulary

*
q
E
t-la-
d
Ba-. Hala-Baba (PN, "the share of Baba")
Ur-
d
Lamar Ur-Lamar (PN)
,._ dub-sar scribe
Commentay
2. The last sign must be a poorly-drawn I-sign.
8. The Akkadian equivalent of b-la is zittu, from the root zizu, "to divide". It is glossd
by the CAD as: "1 . share of an inheritance . . . ". Names of the pattern t-la-DN are com
mon in Sumerian. Similar Akkadian names include tA-LA-i-li and zi-i-DIGIR, "the in
heritance of god".
9. dub means "tablet", and sar means "to write". The formation of the word dub-sar is the
same as that of zabar-dabs= sar.0 is an active participle, and dub is its incorporated object.
The meaning is thus, "he who writes a tablet". The word was borrowed into Akkadian as
tuparu (tanscribed by some as tuparru). It is further discussed in Lesson 21 .
I historic Sumerian, sar means "to write". It i s not known what its original meaning
may have been. In its earliest attestations (para and Tell Abu Salabikh), it does not occur as
an independent verb, but only as part of the nominal compound dub-sar. dub-sar also oc
curs in the bilingual Eblaite texts, but unfortunately without an Eblaite equivalent.
L0880u 20
This is a brick building inscription of Shu-Sin.
Sign-list and vocabulary
1
-ba-Iu
5
-g Habaluge (PN)
Adab Adab (ON)
-

n
ensi
2
ensi (city goveror, loal ruler)
Notes
arad servant
1-ba-Iu5-g The etymology of the name is unsure; it is discussed below.
Adab This is the name both of a city and of a kind of bird. The modern name of the city
is Bismaya. A number of third millennium (and later) texts have been found there.
The etymology of the name is unknown; it is not sure if the word is originally
Sumerian. Lexical lists, and rare syllabic spellings, most often show the initial consonant to
be Idl, but spellings with Irl and Isl also occur. Akkadian spellings also show similar
variation. Such oscillation in spelling shows that the consonant was probably not a simple
I dl, but something more complex, perhaps the I d I phoneme discussed earlier (Lieberman
reconstucts the original name of the city and the bird as lorab/).
There is also some oscillation in the vowels of the name. The canonical spellings
show la-ai, but most earlier spellings seem to show lu-a/. The form ladabl is pre
sumably a result of vocalic assimilation: lu-a/ ) la-al. The Akkadian spellings most often
show lu-a/; that is, they refect the Sumerian pronunc
,
iation of a period when the vocalic
assimilation had not yet taken place.
The name of the city and of the bird is written by what appear to be two cuneiform
signs, the ud-sign followed by the nun-sign. The situation is probably more complex,
however. According to Jacobsen,
Early occurrences show a strong tendency to combine the two later signs V
and NUN into a ligature as if they originally formed but one single larger
sign . . . . The sign is not a ligature but an original pictograph representing a
disc placed on top of a pole or stake . . . . It thus becomes likely that the writing
of the city-name Adab was originally a picture of a symbol, a disc affxed to
a stake for carrying, and since that picture served also to designate the usabu
bird one may assume that the symbol represented an usabu bird and had a
picture of that bird on its disc. Symbols of this kind are well known
( 1 968: 1 01 ).
ensi2 The function of this offcial has been much discussed; it changed over time.
223
22
4
Lesson 20
According to Hallo, the term means "titular head of a city and its dependent territories".
Jacobsen says:
The ttle ensik . . . seems to denote specifcally the ruler of a single major city
with its surrounding lands and villages, whereas both "lord" (en) and "king"
(! ilimply ruler over a region with more than one important city. As for
the origins of the offce, the ensik seems to have been originally the leader of
the seasonal organization of the townspeople for work on the felds:
irrigation, ploughing, and sowing (1957:384 n.71) .
But later, during the Ur lIT period,
The top provincial civil administators, the ensiks, became prper goverors,
entirely dependent on the king, and were moved at will fom one post to
another to minimize the dangers inherent in too stong local ties. Military
affairs were out of their hands entirely (1957: 155).
During the Ur ITI period, the ensi
2
was the highest-ranking civilian authority; the cor
responding military ofcer was the akkana (often translated "military goveror"). Many
of the latter were sons of the king. Occasionally, one and the same person served as both
ensi
2
and akkana. The names of many ensi
2
s of the Ur III period have been preserved.
ensi
2
was loaned into Akkadian, appearing as ii)akku and iakku. It is glossed by
the CAD as: "territorial ruler (of cities, countries, etc)".
The etymology and writing of the word are discussed below.
arad The original form of this sign was the nitab-sign (meaning "man") followed by the
kur-sign (meaning "mountain"). Presumably, the Sumerians derived some of their slaves
frm foreign, mountainous areas. The shape of the arad-sign, however, tends to become
simpler, and even as early as the Ur lIT texts, the nitab-sign and the arad-sign can look quite
similar.
The Akkadian word for "slave" is wardu. This would seem to derive from the verbal
root waradu, meaning "to go down to"; this root occurs in other Semitic languages. There
fore, most scholars assume that the Sumerian word arad was borrowed frm the Akkadian
wardu. A minority view, however, says that wardu has nothing to do with the verbal root
waradu, and that wardu is a borrowing from Sumerian.
A complicating factor is the fact that the reading of the sign is uncertain; some read it
as arad, and others read it as ir. The Sumerian evidence is ambiguous. According to Gelb,
"generally, the form ending in -d is younger than the form ending in -r" ( 1982:86). Thus,
in pre-Sargonic Lagash there occurs: (NIT A1xKUR)-ra-ni, "his slave", but in Sargonic
and Ur III texts there occurs: (NITAIxKUR)-da-ni, "his slave".
Most scholars, however, seem to derive the form in Irl from that in Id/. For
example, Falkenstein assumes a development along the lines of: */ward-al > */urd-a/ >
*/ird-a/ > */irdl > fir!. Gelb, on the other hand, believes that there were originally two dif
ferent words for slave: a native Sumerian word, written II I (ARADxKUR), and also a
borrowed word, written arad (NIT A1xKUR).
There has been much recent discussion about slavery in the Ancient Near East. The
term arad is variously translated as "servant" or "slave". But as Sollberger (and others)
Lesson 20 225
have pointed out, "The usual translation 'slave' is a misnomer because its legal implications
do not ft the Sumero-Akadian social context" (1966: 1 37). In a similar vein, Gelb has
said, "Freedom is relative and the terms for 'slave
'
are quite ambiguous in the Ancient Near
East, as they are in the Classical World, or for that matter anywhere else" (1979a:284).
226
Lesson 20
Text 20
0
I
j

@
lQ4 l
Lesson 20
227
Notes
Line 5: The pad-sign is here split into two parts. This is a common scribal practice with
this sign. It is presumably done for aesthetic reasons; otherwise there would be too much
empty space in the line.
1 :
2:
3:
4:
5:
6:
7:
Transliteration
d
Su-
d
Zuen
ki-aga
2
-
d
En-lil-h
- - - -- -
lugal
d
En-HI-le
ki-fgf
2
-ag
4
-gf!-na
in-pad
lugal-kalag-gf
I I U '
ki
.gf- nms -ma
Transcription Translation
Susin For Shu-Sin,
ki.aga.a.Enlil. a(k) the beloved of Enlil,
lugal Enlil.e the kin_whom Enlil
selecte
ki.aga.0 ag.ani.a in his loving heart,
i .n.pad
lugal. kalaga the mighty king,
lugal. Urim. a(k) the king of Ur,
8: lugil-an-ub-da-limmuy ba lugal.anub.da. (k) limmu. bi.a(k) the king of the four
quarters,
9: digi-ki-g-M-f-ni digir.ki.aga.a.ani. (r) his beloved god -
10: I.-ba-Ius-g Iabaluge Habaluge,
1 1 : ensi
2
ensi. the ensi of Adab,
12: Adab
ki
Adab. (ak)
1 3: arad-da-ne arad.ani.e his servant -
14: -ki-g-M-.-ni e.ki. aga. a.ani. 0 his beloved temple -
1 5: mu-na-du mu.na. (n.)du.0 built.
Commentary
1-7. These lines are identical to lines 3-9 of Text 1 8.
9, 14. Line 9 is written: digi-ki-g-g-f-ni; line 14 is written: -ki-g-g-.-ni. In the pre
vious inscriptions, such expressions were spelled: . . . -g-ni. In Text 17, for instance, lugal
ki-g-g-ni-i! and en-ki-g-g-ni occur; in Text 19 -ki-M-g-ni occurs. Text 20 thus uses
a fuller writing. It is very diffcult to understand the motivation behind such variation in
spelling.
1 1- 12. Presumably a genitive phrase, ensi.Adab. (ak). As in other inscriptions, there is no
graphic expression of the genitive when the second element of a genitive phrase is a
geographical name.
13. Since this is the agentive nominal phrase, there must be an ergative marker .e present.
In other inscriptions, in fact, this line is spelled: arad-da-ni-. In line 1 3, the Ii/ of the
possesive-suffx .ani has contracted into the lel of the ergative case-marker. This
contraction is indicated in the transliteration by: arad-da-ne. However, the ne-sign is the
ni-sign. Some Sumerologists would transliterate line 13 as: arad-da-ni, because they wish
22
8 Lesson 20
to avoid a possible prejudging of the written form. The latter view would regard tans
literations of the type "arad-da-ne" as an attempt to jiggle the script to ft our views of
Sumerian grammar. A similar instance was seen in Text 19, where ma-da-ne- occurred;
others would read this as ma-da-ni-. It is also possible to interret writings of the type
arad-da-ni- as historical or morho-graphemic spellings. It is such ambiguities in the wri
ting system that make it difcult to establish rules governing Sumerian phonology, espe
cially those govering vocalic contraction.
Discussion: structure
The bare-bones structure of this text is:
1 -9 For Shu-Sin
1 0- 1 3 Habaluge
14 his beloved temple
1 5 built.
benefactive
agent
patient
verb
In all the inscriptions in the body of texts seen up to this point, the initial datival phrase
referred to a god or goddess, the deity for whom something was done or built. However, it
is also possible to fnd inscriptions recording actions performed by a subordinate of some
kind, in order to curry favor with the king. Hallo considers these to be royal inscriptions,
because there is a fair amount of latitude in his defnition of royal inscriptions: inscriptions
which are dedicated either "by, or to, or on behalf of the king".
There are not many such inscriptions. Hallo lists only fve building inscriptions of this
type. Curiously, they all come from the reign of Shu-Sin. king.
-Phonology
There are other cases of Idl - Isl alteration in Sumerian. Sometimes this alteration
shows up in different syllabic writings of Sumerian, sometimes it shows up in differing
Akkadian versions of loan-words or of proper names. The goddess of the scribal art, for
instance, variously appears as Nidaba or Nisaba. Some Sumerologists think that this
alternation means that Sumerian had a voiced interdental fricative, IM, which it was
diffcult for the script to represent.
There are several cases of bisyllabic (or polysyllabic) Sumerian words which show
vowels of only one quality, but whose Akkadian equivalent shows vowels of two different
qualities. For example, there is a tree usually spelled za-ba-Iam in Sumerian. The word
appears in Akkadian as supalu. Most probably, this is a pre-Sumerian substate word,
which passed into Sumerian. It then passed into Akkadian, presumably through Sumerian
(or conceivably by a different route). The Akkadian preserves the older vocalization; the
Sumerian form shows that at some time Sumerian underwent a rather extensive process of
vocalic assimilation, a process of which only traces can be seen. Assuming that Akkadian
borrowed the word fom Sumerian, it must have borrowed the word before the Sumerian
vocalic assimilation had taken place. This same process was seen in Lesson 1 1 , where the
word zabar was discussed; the pronunciation Izabarl is the result of vocalic assimilation
from some earlier form of the word like Isibar/; Akkadian again preserves the more
Lesson 20
229
original vocalization.
-ensi
2
There is no unanimity among scholars about the etymology and the writing of the ter
ensi
2
. (The latest summary is in Dunham 1986: 5 1 -52.)
The word probably had a Ik/-Auslaut. When it is followed by a vocalic ending of
some kind, a Ikl usually appears; note also the Akkadian ii'akku. It has frequently been
proposed that the Ikl is the genitive marker; the word may originally have been a genitive
phrase, en. si. (k), "the lord of si". Unfortunately, it is not sure what si means here.
Jacobsen thinks that si means "arable land"; the title would originally have meant "manager
of the arable lands". But it has also been argued that the word has a pre-Sumerian substate
etymology, and that the interpretation "lord of si" is a Sumerian folk-etymology.
Some Sumerologists now transliterate this word as en
S
-si. This interpretation is
partially based on the understanding of the word as a genitive construction, even though the
spelling with en
S
(instead of simple en) is weird.
The problem of understanding the writing is complicated by the fact that although
normally written PA-TE-SI, the word is also occasionally written PA-SI, and even just PA;
other bizarre spellings are attested. These writings make it difcult to understand how the
cuneiform signs are meant to refl ect the language (should they be regarded as
abbreviations?). Such writings cause numerous problems in transliteration.
-Genitive
One of the pieces of evidence which indicates that the final Ikl of the genitive
morpheme was actually pronounced in spoken Sumerian is the fact that it occasionally turns
up in words loaned into Akkadian. For instance, the Sumerian god Amar-Utu, "young bull
of Utu", appears in Akkadian as Marduk, presumably from amar. utu. (k) If the Sumerian
ensi
2
is indeed a genitive phrase, then the Akkadian ii)akku also shows the genitive mar
ker Ik/. It has, however, also been argued that "Marduk" is not of Sumerian or of Ak
kadian origin, and has nothing to do with "Amar-Utu".
-Proper names
The personal name "Habaluge" occurring in this inscription is presumably of
Sumerian origin, although the exact etymology is not known. When a name is of uncertain
etymology, it is ofen difcult to determine the precise reading of the signs. For example,
the third sign in this name can be read Iu
S
'
lul, lab, nar, etc. The fourth sign is also gra
phically ambiguous; several different signs share approximately this same shape. For
example, the -sign and the ke4 -sign, even though distinct signs, look very similar in sev
eral of these royal inscriptions. And if it is the ke4 -sign, it has several different readings:
ke4
'
, lil, etc.
One way such ambiguities can be resolved is to identify different spellings of the same
name. For example, this particular name is not uncommon in Sumerian texts. It is most
commonly tansliterated as: I-ba-lu
S
-ke4
'
with the proviso that the reading of the third
and fourth signs is not certain. There are also spellings where the third sign appears as lu,
230 Lesson 20
and not as the lus -Iul-Iab-nar-sign: l-ba-Iu-g and l-ba-Iu-g
1 8
. This shows that the
third sign is probably to be understood as Ilu/, and so to be read as lu, and not as lul, lab,
or nar. The fourth sign is spelled as , g, and g
1 8
: l-ba-Ius -
'
l-ba-Iu-, and l
ba-Iu-g
1 8
. This shows that the last sign is a g/ke4 -sign, and not the graphically similar
sign; it also shows that the reading was probably Ige/. Therefore, the most likely reading
ofthe name i Text 20 is: l-ba-Ius-g.
The latest discussion of this name is by Steinkeller ( 1 984:9); he reads the third sign as
lugx
'
that is, with a Ig/-Auslaut. He believes that the meaning of the name "cannot be
gauged with confdence", but it is undoubtedly a Sumerian verbal form with the modal
prefx b; it may mean "May-he-pasture/take-care-of
'
.
Lesson 21
The three texts in this Lesson do not offer very much new in the way of grammar, but
they illustate common types of royal inscriptions. The frst is a weight of Shu-Sin. The
next two are seals of rbbi-Sin, the son and successor of Shu-Sin; he was the last ruler of
the fDynasty.
Sign-list ad vocabulary
W "5"

ma-na "mina" (measure of weight, about 33 grams)


&i-oa true. correct; standard. certifed
Notes
5 When Sumerian numerals are used strictly for counting, they are normally tansliterated
by Arabic numerals. The word for "fve" was iA.
ma-Da hs etymology and value are discussed blow.
gi-na This is a loan from Akkadian klnu.
Text 21 a
231
232
Transliteration
1 :
2:
3:
4:
5:
ma-na g-na
d
Su-
d
Zuen
lugal-kalag-g
I I U '
ki
..w- nms -ma
lugr-an-ub-da-limmu
T
ba
Commentay
Lesson 21
Transcription Translation
5 mana gina 5 standard minas.
Susin Shu-Sin,
lugal.kalaga the mighty king,
lugal.Urim.a(k) the king ofUr,
lugal.anub.da.(k) limmu. bLa(k) the king of the four
quarters.
1. In simple enumerations, the numeral is followed by a singular, not a plural. In many
languages, the noun appearing after a numeral assumes special forms, in number or in case.
Sumerian shows no outward evidence of this; in general, agglutnative languages show no
special forms after numerals.
2-5. There is no verb form.
Discussion: weights
This is a typical weight inscription. Unlike the weight inscription seen in Lesson 10,
this one actually gives the weight. It is diffcult to say exactly what gi-na means in such
contexts; it is usually understood as "standard", or perhaps "certifed". One of the more
important functions of Mesopotamian rulers (throughout all periods of history) was the
regulation of the system of weights and measures, but not much is known about how such
weights were actually managed by the crown. Gadd points out that Shulgi "rearranged the
calendar, set up a bureau of standards, and issued accurate weights which were preserved
and imitated to the latest days of Babylonian history" ( 1 971 : 61 8). In the prologue to Ur
Nammu
'
s Law Code, there is a section referring to the "standardization" of the mina; this is
usually interpreted as referring to some kind of reform of the royal weight system. Re
cently, Irving Finkel has published a text dated to Amar-Sin
'
s fst year, which is a "receipt
for two differing sets of weights":
The implication of the text . . . is that an offcial issue of correct weights was
made at the beginning of Amar-Sin
'
s reign, and that this document reflects a
deliberate attempt to ensure that government offces were using uniform
weights ( 1987: 192- 1 93).
-Metology
All the Classical Semitic languages except Ethiopic have a verbal root *mnw/y,
meaning "to count, to reckon". Therefore, Sumerian ma-na is probably a loan from the
Akkadian manu (which is also the ultimate source for the English word "mina").
The value of the ma-na and the manu varied to some degree from time to time and
from place to place (Powell refers to "a multiplicity of standards which defes reduction to
one or more 'common
'
norms"). In Sumer proper, the most common value of the ma-na
was about 505 grams. In Mesopotamia, the manu was the same. But in most of Syria, the
Lesson 21 233
manu was a little less, from 470 to 480 grams. At Ebla, it was also about 470 grams. (The
theoretical value of weights is determined simply by averaging out the actual weights of
weights inscribed with their values.) Since this particular weight is a . ma-na weight, it
should weigh about 2525 grams. It actually weighs a little less, 251 1 grams.
The Sumerian ma-na was divided into 60 gi n; the Akkadian manu into 60 iqlu. 60
ma-na formed a gun; 60 manu formed a biltu:
- Numbers
60 ma-na
60 gi n
1 biltu 60 manu
1 manu 60 iqlu
The pronunciation of the numeral for "fve" as fia/ is known from mathematical
cuneiform texts, where it is occasionally spelled out. The Ebla school-text mentioned in
Lesson 1 0, which spells out the Sumerian numerals from one to ten, simply gives i for "5".
Unfortunately, not enough is known yet about the nature of the Eblaite syllabary to say
what values the i-sign could have had at Ebla; however, there is some evidence to show that
one of its values was /ya/. Pettinato, in fact (198 1 : 143) reads the sign on TM.75. G.2198
as i, but this is perhaps somewhat adventuresome.
>
It is also signifcant that the i-sign is composed of fve strokes: ;:.

Sign-list and vocabulary
%- Nanibgal Nanibgal (DN)
+ I-bi -
d
Zuen Ibbi-Sin (PN)
n_ Da-da Dada (PN)
*
*
Ur-
d
Nanibgal Ur-Nanibgal (PN)
zu your
Notes
I-bi-
d
Zuen The name is good Akkadian, meaning "Sin has called"; ibbi is the preterite
from nabu, "to call".
Da-da The etymology of the name is unsure; it is discussed further below.
Nanibgal Very little is known about this deity. The reading of the name is somewhat un
certain, as is the etymology. The cuneiform character appears to be the an-sign followed by
the nisaba-sign. Therefore, the name is sometimes transliterated as AN.NISABA, or as
234 Lesson 21
DIGIR.NISABA (Nanibgal is not infrequently mentioned alongside Nisaba). In the writing
in Text 21 b, the determinative and the an-component of the sign are written on top of each
other, instead of following each other. This is for the sake of graphic symmetry; other
times the two components are written after each other.
0Xl Z1 D
Notes
The name of Dada
'
s father, occurring in line 1 1 of this text, illustrates the diffculties of
working from autographs. The autograph seems to show the name as Ur-tg
S
. tg
S
(or
a
6
) means "good". The name would then be a variant of Ur-ags-g, a relatively common
personal name meaning "the good man". However, the photograph of the text is more
ambiguous. The sign in question is actually somewhat damaged, and the remaining traces
can be made to ft either ags or
d
Nanibgil.
The reason for preferring the reading
d
Nanibgal over tg
S
is because other seal im
pressions have been preserved of this same individual, Dada, in which the sign for the
name of his father is more distinct, and in these the sign is clearly Nanibgal.
One cannot always accept a modern-day editor
'
s transliteration of a text. An autograph
caries more evidential value, but even then cannot always be accepted at face value. This is
especially true for autographs which were drawn when knowledge of Sumerian was weak.
Better than a transliteration or an autograph is a photograph. However, for many published
texts no photographs are available, and for others the photographs are reprduced in such
poor quality that they are almost useless. For any signifcant passage, there is no alterative
to a close examination of the original cuneiform document.
1 :
2:
3:
4:
5:
6:
7:
8:
Transliteration
d
!-bi -
d
Zuen
digi-kalam-ma
lugil-kalag-
I I U
ki
-
Yg
-
nm
5
-ma
lugal-an-ub-
da-limmu
2
-ba
Da-da
9: ensi
2
10: Nibru
ki
1 1 : dumu Ur-
d
Nanibgil
12: ensi
2
1 3: Nibru
kC
ka
14: arad-zu
Comentary
Lesson 21
Transcription
Ibbisin
digir.kalam.a(k)
lugal.kalaga
lugal. Urim.a(k)
Translation
Oh Ibbi-Sin,
god of the land,
mighty king,
king of Ur,
2
35
lugal.anub.da.(k) limmu.bi.a(k) king of the four quarters -
Dada Dada,
ensi the ensi of Nippur,
Nibru. (k)
dumu. Urnanibgal the son of Ur-Nanibgal
ensi the ensi of Nippur -
Nibru.k.a(k)
arad.zu is your servant.
1 . It is usually assumed that in seal inscriptions the initial nominal phrase contains the
name of the king in a vocative. The vocative normally has no formal marking in Sumerian.
Another example occurs in Text 22. There are a few cases where the vocative is marked by
.e. This is presumably an extension in use of the locative-terminative.
The essence of this seal is: "Oh Ibbi-Sin, Dada is your servant".
13. A double genitive is contained here:
Dada,
ensi.Nibru.(k)
dumu. Urnanibgal
ensi.Nibru..a(k)
14. zu is the second person possessive-suffx, "your". It is diffcult to say whether this is
a nominal sentence without the copula ("Dada is your servant"), or an appositive ("Dada,
your servant").
Discussion: possessive-suffxes
The possessive-suffxes for the singular are:
frst person
second
third animate
inanimate
- g!
lQ
(mu, g!
IQ
)
- zu
- .! -ni
- bi
23
Icn2!
The initial consonant of the marker for the frstperson singular is the velar nasal
discussed under Phonology. conventionally tansliterated by Sumeroiogists as g. There
fore, glO is the most up-ta-date reading of this sign in this usage. However, it is also
tansliterated as gylO' and (especially in older works) as mu.
-Proper names
The name Da-da ocurs very frequently in cuneiform texts. It is not very easy to de
termine the etymology of a name of such a simple stucture, what appears to be a redup
licated CV syllable. The name has variously been considered to be Sumerian, Akkadian, o
"other", In an article on "Ethnicity and Onomastics in Sargonic Mesopotamia" (lV82),
Benjamin Foster divided personal names into four groups: Sumerian; Akkadian; Redup
licated; Unsure. He purposely omitted the name gfrom discussion, because of the
diffculty in detennining its etymology.
Such reduplicated personal names are sometimes referd to by the Gennan tenn
"Lallnamen" or "LallwOrter" ; another example is Du-du.
-Seals
Seals were used by ofcials in Mesopotamia (and elsewhere) to stamp their "seal of
apprval" upon douments of all kinds. The act of sealing could perfon several functions,
such as acknowledgment, authorization, guarantee, etc. Mesopotamian seals usually
consist of two components: a pictorial scene of some kind, and a short inscription. Many
seals have only a pictorial scene, and lack an inscription. There are only a few seals which
have just an inscription, but lack a pictorial scene. The pictorial representation on the seal
frequently has a mythological signifcance, as in the example given below. Or, it may be a
"presentation scene", usually thought to represent the possessor of the seal paying homage
to his ruler, the king. The following is a picture of a (non-royal) seal, fom the Old
Akkadian perio. (This partiCUlar seal has often been reproduced.) The inscription reads:

mm-sar, "Adda the scribe". The photograph is actually of the impression of the seal,
not the seal itself. This is because the text on seals is inscribed in reverse (that is, mirror
imaged), so that when impressed, the text comes out in the right direction.
'
f

'
_
'
j
)
\
;
I
'
A
' 1..
Lesson 21 237
It is not easy to detenine what language such a short inscription is written in. It may
be perfectly good Sumerian. However, according to the criteria used by art historians, the
seal in the photograph dates to the Old Akkadian period. Therefore, the language is
probably Akkadian, and dub-sar should be read as a logogram for tuparru.
- Scribes
Scribes occupied a central rle in all of Mesopotamian civilization. Many modem
scholars have written about them, but there is still much that is not known; for example, to
what degree was their job hereditary; what was the extent of their infuence at the royal
court; did they have non-scribal work at the court or elsewhere, etc. Writing on the "social
position of Neo-Babylonian scribes", Muhammad Dandameyev has said:
Our infonation about the social position of the Mesopotamian scribe, his
activity as bureaucrat and in the service of the community for recording of
contracts is very scanty. We have no direct data on the economic situation
and the social origin of scribes. We also do not know if the scribal
profession was the chief source of income or if the scribes were busy with
their craft along with handicraft, tilling ofland and so on ( 1 982:35).
C. B. F. Walker points out that
The scribes, like any craftsmen, had to undergo training, and having
completed their training and become entitled to call themselves dubsar
"scribe", they were members of a privileged elite who might look with
contempt on their fellow citizens ( 1987:33).
Specifcally discussing seal-practice in the Ur III period, Steinkeller says:
The ten dub-sar, apart from its basic meaning "scribe" is an honorifc title
which merely indicates the graduation of the individual in question from a
scribal school. . . . It is tempting to speculate that the "dub-sar seal" was a
kind of "diploma", which may have been presented to a graduate of a scribal
school at the conclusion of his studies. The possession of such a seal would
have constituted proof that its owner was eligible and entitled to be employed
in the state or temple administrative apparatus or to sell his services to private
individuals ( 1 977:47-48).
In a similar vein, Veenhof mentions the role of scribes
in the administration and their position in society, which may range from
that of a simple clerk or a paid letter writer on the market to that of a chief
accountant or secretary of a chancery or king ( 1 986:21 ).
The rather automatic translation of dub-sar as "scribe" paints a rather simplistic picture.
Michalowski says that "In Ur III times dub-sar was a general term for low and middle level
bureaucrats" ( 1 987: 62).
And to quote Walker again,
[Most scribes,] afer all their technical training, spent their lives writing lists
of deliveries of sheep or issues of barley rations and occasionally taking a
letter by dictation. The more successful scribes would end up as senior
administrators in the state bureaucracy, but most of their colleagues would
238 Lesson 21
have been happy simply with their status as educated men and the
knowledge that their taining guaranteed them employment ( 1987: 39).
In a thought-provoking article about what we don't know about Ur II society, Soll-
berger asks:
We know roughly what the prfessional scribe
'
s jobs consisted of, but how
did he work, and where did he work, and how did he make himself known
as a professional scribe and his services available? And there is of course
the nagging question which is usually politely glossed over: where did the
scribes get the enormous amount of clay they needed? Were there clay
stationers? Did one have to buy clay or did one just go to the canal bank and
help oneself . . . ? ( 1972: 1 88).
-Engravers
When dealing with "monumental" inscriptions, it is necessary to distinguish between
"scribe" and "engravers" (or, "lapidaries"). The latter were the persons who actually
chiselled the inscriptions into the stone. They were not always literate, but simply copied a
design or plan, which may have been drawn onto the stone. Presumably, the engraver
worked under the supervision of a scribe. In other cases, the scribe and the engraver may
have been the same person.
The standard word for "engraver" was zadim. This word apparently derives fom za,
"stone", and dim, "to fashion"; dim is an active participle, and za is an incorporated direct
object.
I the case of seals, the situation was probably a little more complicated. The
inscription and the pictorial scene were sometimes engraved by different individuals. W.
G. Lambert (discussing seals of the Cassite period) asks:
-Seals
A basic question which needs answering is, who carved these inscriptions?
Did one man carve both glyptic and inscription, or were separate crafsmen
employed for the artistic and scribal parts? In some cases it is clear that the
glyptic was carved frst, because not enough room was left for the
inscription, so that the last line had to be spread out among the glyptic. But
in other cases where the inscription covers virtually the whole area, and the
glyptic is reduced to a rw of insects for example, then one may suspect that
the inscription was carved frst and the glyptic was a second thought, serving
merely in fugam vacui. ... One may wonder if two quite separate guilds of
craftsmen were in existence, and such a division of labour seems very
probable in the contemporary boundary stones. ... On general grounds too
such a differentiation is likely, since the artist and the scribe needed very
different training. Yet one need not suppose that this demarcation was al
ways completely enforced ( 1975: 220).
Many seals from the Ancient Near East have been preserved. Even more common
than the seals themselves are seal impressions, that is, the impression of a seal upon a
Lesson 21 239
cuneifon doument. Text 21 b is a seal-impression found on a record of official ap
pointments of individuals to sundry govermental ofces.
Seal impressions are necessarily rather small, being squeezed onto a small seal. This
accounts for some of the odd division of names and epithets put onto more than one line.
This small size often makes it diffcult to read autographs or photographs of seal
impressions, even if reproduced full-size. Occasionally, the script used on the seals is
archaicizing, compounding the problems of reading.
When publishing editions of cuneifon texts, Sumerlogists do not usually present a
drawing of a complete seal as it appears on a document; they will usually only reproduce
the impression left by the inscription. This is primarily because of the mechanical effort it
takes to adequately reproduce (and even just to describe) the scene carved on the seal, and
linguists are not artists or art-historians. Sometmes the seal impressions will be briefy
described, and occasionally published in a separate volume, distinct from the texts
themselves.
The practice of only reproducing the inscription, and not the pictorial scene, is
unfortunate. While linguists may only be interested in the inscription, art historians,
anthropologists, and historians, among others, are just as interested in the scene itself. Seal
impressions are also important for scholars studying groups of documents, and archival
relationships.
-arad-zu seals
Texts 21 b and 21 c may be called ryal seal-impressions, again following Hallo
'
s
defnition of "royal" as: "by, or to, or on behalf of the king". Many non-royal seals and
seal-impressions (such as that in the photo above) have also been preserved.
There are two principal types of Ur III seals. Text 21 b is an "arad-zu" seal. This type,
which is very common, has a specific structure. First, there is the name of the ruling
monarch. This i s a nominal phrase, in the vocative. Second, there is the name of an
offcial, with various epithets or flial relationships as appositives. Third, the ten "your
servant" concludes the seal.
The usual interpretation of arad-zu seals is that d offcial had it cut, out of homage or
respect for the king. However, it has been speculated by Richard Zettler that "the fow of
these seals was fom king to offcial and not from offcial to king" ( 1 977:33).
A few seal-inscriptions have been preserved where the last line reads arad-ni, not
arad-zu. Presumably arad-ni is for arad.ani, "his servant". (This is another instance of the
overhanging vowel problem.) Not enough is known about such seals to characterize them
differently fom the arad-zu seals.
Although arad-zu seals have traditionally been classifed as "Sumerian", with a
Sumerian inscription, it has been prposed to read seals of this type frm the Akkadian
period as Akkadian. That is, arad-zu is to be read as ARAD-su, for warassu, "his servant".
(This is the usual Akkadian fon, resulting from the regular assimilation of */dU ) Iss/:
*/warad-ul ) Iwarassu/.) The rest of the text would then be understood as logograms, to
be read in Akkadian.
The reason for this possibility is because of a seal where the last word is written arad-
24
0 Lesson 21
za, presumably ARAD-sa, for Iwarassa/, "her servant", and another seal written geme
2
-za,
possibly GEM
2
-sa, for lamassa/, "her servant (fer)
,
'
.
While this may be true for seals of the Akkadian period, it is hard to say whether it
might be tue for Ur III (and other) seals. There is really no evidence to decide one way or
the other, and so the question is still open. Barring explicit evidence to the contrary, it is
probably best to assume that for the Ur III period, at least, the text is Sumerian.
-Seals
Both Text 2l b and 2l c may be called functional, in the sense discussed in Lesson 4.
There are also a certain number of votive seals. These are known primarily from the seals
themselves; only a few of the actual impressions are preserved. Text 22, below, is such a
votive cylinder seal.

Sign-list and vocabulary
G:1 Sag-
d
Nanna-zu
Sagnannazu (PN)
sa (kind of priest)
r
ba to give as a gift
Notes
Sag-
d
Nanna-zu sag means "head", but can also have the meaning "slave". zu is a verbal
root meaning "to know". Names of this type are usually interpreted as some kind of
reduced relative or partcipial clause, "the slave who knows Nanna", or "the slave whom
Nanna knows".
saga This priest was high up in the temple hierarchy, although very little is known of his
priestly duties. He seems to have been mostly concerned with running the administrative
side ofthe temple. While the conventional tanslation is "sang-priest", Gelb points out that
"In his capacity as the head of a household, the word san may be interpreted as the chief
administrator of a temple household" (1979b: 1 6); Snell tanslates the title as "economic
director of a temple". The same cuneiform sign, in fact, can be read as ita5
'
"to count"
(Akkadian manu); Aage Westenholz, among others, says that the sign "depicts an abacus or
a countingboard" ( 1 985:296).
Transliteration
1 :
2:
3:
4:
5:
6:
7:
8:
d
I-bi-
d
Zuen
lugil-kalag-r
I I U
'
ki
..g- nm
S
-ma
lugal-an-ub-da-limmu
T
- oa-e4
-
Sag-
d
Nanna-zu
sal_
d
En-li l-hi
- -
-
-
arad-da-ni-i
in-na-ba
Commentar
Lesson 21 241
Text 21 c
Transcription Translation
Ibbisin Ibbi-Sin,
lugal.kalaga the mighty king,
lugal.Urim.a(k) the king ofUr,
lugal.anub.da.(k) limmu.bi.ak.e the king of the four
quarters -
Sagnannazu to Sagnannazu,
saga.Enlil.a(k) the sangf-priest of Enlil,
arad.ani.r his servant -
Lna.(n. )ba gave (this seal).
8. The verb form is almost always written this way. The in-sign contains the conjugation
prefx .i and the initial I nl of the datival dimensional-prefx.
Discussion: arad-da-ni-ir seals
This type of seal is known as an arad-da-ni-ir seal, or as an in-na-ba seal. Its most
common structure is: First, the name of the king, with the ergative case-marker; second, the
name of an offcial, with various epithets; third, the appositive arad. ani, with the dative
case-marker; fourth, the verb form in-na-ba. This results in the datival noun phrase
following the subject, and immediately preceding the verb. In most of the texts seen earlier,
the datival noun phrase occured at the beginning of the text.
The understood direct object in this type of seal is the seal itself. The king gave such
seals to his offcials (and family members), presumably as a reward for some kind of
service. For this reason, they are sometimes referred to as "presentation seals". It is also
possible that the king gave the seal to an offcial upon his appointment to an offce. Fewer
such seals are preserved than arad-zu seals.
arad-zu seals are found under all fve kings of the Ur III Dynasty. arad-da-ni-ir seals,
on the other hand, are only attested for the reigns of Shu-Sin and Ibbi-Sin. This may (or
may not) be due to accidents of preservation. It is also possible that there was some kind of
change in administative practice.
242 Lesson 21
- History
Shu-Sin apparently died a natural death. He was succeeded by his son, Ibbi-Sin.
Early in the latter
'
s reign, the easter territories under the control of Ur broke away, then
other parts of the empire began to fall away. For most of his reign, Ibbi-Sin
'
s control
extended no further than the city of Ur itself. The economy collapsed, and a vicious
infationary spiral ensued. Very little is known of the details of the twenty-four or twenty
fve years of Ibbi-Sin's "reign". Jacobsen has said:
How an empire like that of the Third Dynasty of Ur - to judge by our
sources the most effciently organized stucture of its kind before Assyrian
times - could so quickly and so completely collapse without pressure fom
any enemy state or states of comparable magnitude is really quite puzzling
( 1953: 173).
Jacobsen wrote this over thirty years ago, but his puzzlement still largely stands;
Steinkeller says that "the phenomenal rise of this empire was matched only by the sudden
ness and completeness of its demise; in less than a century after its creation, no trace of it
remained" ( 1987b: 19).
Finally, there came invasions by the Amorites, against whom Shu-Sin had built the
wall mentoned in Text 19, and the Elamites. The Elamites, aided by a somewhat obscure
group of people from the Zagros mountains known as the "Su" or "Sua" (recently
identifed by Steinkeller [ 1988] with Shimaski in Iran) sacked Ur, then withdrew back to
Elam, carying Ibbi-Sin back with them; he died in Anshan. Gadd says: "Ibbi-Sin became
the typical fgure of an ill-starred king, remembered only for his captivity and death in a
strange land" ( 1971 :617).
During their sack, the Elamites destroyed every temple standing in Ur, and all of its
fortifcations. One of the most well-known Sumerian literary compositions is a long poem
entitled "The Lament over the Destruction of Ur", which bemoans its destruction.
This was not the only time that Ur was sacked; Samsu-Iluna of Babylon also levelled
the city, in 1740 BC. The year-date for the eleventh year of his rule is: "The year in which,
at the pleasure of Anu and Enlil, he destoyed the walls of Ur and Uruk". Woolley
describes the destuction:
The ruins bear eloquent testimony to the thoroughness of that destruction.
The fortifcations were dismantled - this indeed one might expect; every
temple that we found had been plundered, cast down, and burned; every
house had been consumed with fre; the whole of the great city ceased to
exist ( 1 982: 214).
However, the city was rebuilt almost as ofen as it was sacked. Most kings of the
Isin-Larsa period rebuilt old temples and built new ones. Such construction took place
right through the Neo-Babylonian period. Although Ur never regained the political
importance it enjoyed under the Ur III Dynasty, there were times when it must still have
been an imposing city.
Ur was occupied - at least, to some degree - into the Persian period. It began to
completely fade out about the fourth century BC, through the effects of changing trade
L6 n2! 24
patterns. the shifting of the course of the Euphrates and the concomitant loss of agriculture,
etc.
Text 21d
supplementary
Notes
Line 8, In Text 21a, gi-na (a brrowing fom Akadian) was used as an adjective.
However. it can also D used as a verb: mU-!-gi-in. It is usually tanslated as "he stan
dardized". or perhaps here "he certfed". As discussed abve, the exact signifcance of the
tenn is not clear.
The actual weight of this weight is 2478 grams.
Lesson 22
This is a votive cylinder seal. made of limestone. dedicated to the tife of Shulgi.
Sign-list and vocabular
< r < Jl n Mes-Iam-ta--!
Meslamtaea (DN, mase)
8 :i Ki-Iul-Ia Kilula (PN)
L: r @-g-? Urba-x (PN)
-( LagaS <SIR.BUR.LA) Lagash (ON)
,i
a, strength
245
246 Lesson ::
_ f gg-za throne
_ gg-za-hi (kind of offi cial)
_ kiib cylinder-seal
d
T>-
_
getug
3
ear; intelligence
J
nig (n1 ) thing
|
hi to hold, to lift, to carry
:
ag
5
(6)
to be good, pleasant, nice
i
Notes
Mes-lam-ta-e-a God of the Netherworld, apparently the same as Nergal. Mes-lam (also
transliterated as MM-lam) is the name of the temple of Nergal in Kutha. Its meaning is
unknown (although
gi
mes is well known as a type of tee). ta is the ablative case-marker.
r is a verbal root, meaning "to go out". is the nominalizer. Thus, the name means
something like "he who goes out of the Meslam-temple".
Meslamtaea himself does not occur fequently in texts, but Nergal was widely
worshipped in Mesopotamia. He shared the rule of the Netherworld with Queen
Ereshkigal.
Ki-lul-la This not uncommon name is presumably Sumerian, although the etymology is
unsure. It is variously spelled as: Ki-lul-la, Ki-lul-hi, and Ki-lul-hl-.
Ur-ba-? The reading of the third sign is uncertain. Both the autograph and the photo
show a sign which looks closely like the last sign in the inscription, which must be bi. The
reading bi for this sign is also the reading preferred by Gadd, who says that upon a "fresh
examination" of the seal, "the engraver certainly traced, and doubtless intended, the same
sign as in the last place of the whole inscription", that is, bi. However, there seem to be no
parallels to a putative name Ur-ba-bi. (Gadd speculates that Ba-bi may be a variant of the
name of the goddess seen in Lesson 1 1 , spelled as
d
Ba-., which earlier Sumerologists
read as
d
Ba-..)
Most Sumerologists have interpreted the sign as a poorly drawn -sign (if not a
simple mistake). The g-sign here would be read as gar. Ba-gar is the name of a well
attested temple of Ningirsu at Lagash (the etymology of Ba -gar is unsure). The personal
name would then mean, "the man of (the temple) Bagara". This name is attested elsewhere.
Also, other names composed of Ur and the name of a temple are attested.
Lesson ::
24
7
Lagd "Lagash" i s used in two senses in Sumerian (and in English). Stictly speaking, it
refers to a city proper (whose moder name is el-Hibbe). But it can also be used to refer to
the territory controlled from the city of Lagash. This larger entity i s sometimes referred to
as "Lagash-state", as opposed to "Lagash-city". This included Lagash-city; the adjacent
Girsu (whose moder Arabic name is Telloh, meaning "mound of the tablet"), and Nina
(moder Surghul). Because of this ambiguity, the term "Lagash" is sometimes used when
referring to Girsu.
This was the site of the frst important excavation of a Sumerian tell, begun by the
French in 1 877. Thousands of tablets were found, including a number of royal inscriptions
from the First Dynasty of Lagash, and fom the time of Gudea. These are one of the prime
sources of Sumerian for the period.
The etymology of the name is unknown. It is not known how the three signs SIR
BUR-LA came to represent the name of this city. One would guess that the la-component
is some kind of phonetic complement. The name is, in fact, occasionally written LA-BUR
SIR.
I This originally meant "d", and then "strength". In this latter sense, its Akkadian
equivalent is emlq!, glossed by the CAD as: "1 . strength (in physical sense as localized in
the d), . . . ". J-zid-da is something like "the effective d". It can be approximately
rendered in English as "right arm" or "right-hand-man".
In Text 17, the expression digi-zid occurred, translated as "the effective god". There,
zid appeared without the nominalizer -. But here, the same adjective occurs with the
nominalizer -: zid-da.
gu-za-ll g!-za means "throne". hi is a verbal root meaning "to hold, to lift, to cary".
Here, hi is an active participle, and g!-za is its incorporated object: "he who holds the
throne", or "throne-holder".
g-za-la was borrowed into Akkadian as guzalU, explained by the CAD as: "an off
cial, lit. chair-bearer, originally a servant carying a chair after his master".
g-za is thought to be the ultimate source of the word for "chair, throne" appearing in
many Semitic languages: Akkadian kussu, Hebrew kisse, Arabic kursTy, etc. But since the
patter of the Sumerian form is a little stange, it has more recently been speculated that g
za is a borrowing fom Akkadian. This situation illustrates the difculty of evaluating the
evidence of loan words; it is not always easy to establish the direction of a borrowing.
kBib It has been speculated that this is a pre-Sumerian substrate word.
getug
3
This can be spelled in a variety of ways. The simplest was originally the picture
of a donkey
'
s ear ( <T>-), tansliterated as getug. It can also be written with the signs
and ilfunctioning as phonetic complements; the combination of these three individual
signs can appear in varying orders:
ge-tug
getug ( getug
2
)
ge
getug
tug
( getug
3
)
Although it is clear (now) that ge and il in this particular use are phonetic
248 Lesson ::
complements, they are not usually transliterated as such. Since the simplest (at least,
phonetically) reading of the getug-sign is m, Sumerologists often refer to this sign as the
"m-sign", and tansliterate this word as: ge-tUg-m, or (as here) : ge-il -tug.
In its concrete meaning as "ear", the Akadian equivalent of getug is uznu. However,
it is also equated with basTsu, glossed by the CAD as: "(1) aperture of the ear, ear, (2) (fa
culty of) hearing, (3) understanding". Jacobsen says that "the Mesopotarans believed the
ear, not the brain, to be the seat of intelligence".
ags Variously transliterated as: ag
5
'
a
6
'
sag
9
'
sa
6
'
sig
6
, etc. The most comon Aka
dian equivalent is damlqy. The verbal adjective damqy is glossed by the CAD as: "1 .
good, fine, pleasant, 2. beautiful, handsome, 3. of good family, well-to-do, 4. expert,
well-tained, 5. of good quality, in good condition, 6. gracious, favorable, 7. propitious,
8. effective, 9. canonical".
r.--::
0Xl 22
4

J
=
"
J| | |

'
(

-'
,
+'\
z
H "
=
1^


"

||

!
JJ |

\~|
+
<

` . -

249
25
0 Lesson ::
Notes
Line 9. Assyriologists use a superscript exclamation mark (
!
) to indicate that the cuneiform
sign so marked in tansliteration is written in a deviant or aberrant fashion on the tablet.
For example, the proper name in this text might b tansliterated as: Ur-ba-gar
!
. This
transliteration means that in the opinion of the modern editor, the sign is a poorly drawn
gar-sign.
Another convention is to employ the superscript exclamation point to mark the correct
value in transliteration (that is, what the editor thinks it should be), and to follow this with
"what is actually written", in caps within parentheses. Thus, a tansliteration such as: Ur
ba-gar
!
(BI) would mean that in the opinion of the editor, the sign drawn on the tablet is
bi, but the editor believes that this bi-sign is a mistake for gar -the scribe made an error.
As can b imagined, these conventons can b misleading or confusing. They illustrate
the importance of working directly from the texts, from photographs when possible, or
from autographs, and not just from tansliterations.
Line 10. Because of damage to the cylinder seal, this line is somewhat diffcult to read. Of
the getug
3
-sign, the ge-component _,is clear. The I (or getug)-component . ,is
virtually completely effaced. Of the !g-component ( _), only the bottom shows. In
addition, the following ni g-sign , ,is scrunched up.
1 :
2:
3:
4:
5:
6:
7:
8:
9:
Transli teration
d
es-lam-ta--g
lugal

-zid-da
Law
ki
-ke4
nam-.-i
d
Sul-gi nitab-kalag-gg
lugil-Urim
5
ki
-ma-ka-e
- - ---
Ki-lul-la g!-za-la
dumu Ur-ba-?-ke4
mu-na-dim kiib-ba
10: lugil-gu
l
O
Retug
3
-nig-ag
5
-
gg-Ia-ne
1 1 : gg-an-.-i
12: mu-bi
Comenta
Transcription Translation
eslamtaea For eslamtaea,
lugal a.zid.a the king, the right-hand man
of
Laga. (a)k.e Lagash -
nam.til for the life of
Sulgi nitab.kalaga Shulgi, the mighty man,
lugal. U rim.ak.a(k).e the king of Ur -
Kilula guzala Kilula, the guzala-offcial,
dumu Urba? ak.e the son of Ur-ba? -
mu.na.(n
'
dim
kiib.bi.a k)
fashioned (this). The name
of this cylinder seal is:
lugal.gu
.
getug.nig. ag.
ak. am.e
"0 my king, let me kee
l
him
alive at his ear of favor .
ga.(L)n.til
mu.bi
3. Laga.ak.e, written Laga
ki
-ke4
'
The followers of the Falkenstein school would read
the frst sign as Lagaa.
The .e at the end is the marker of the locative-terminative case. In all the previous
r.--::
251
instances of indirect objects and benefactves, the nominal phrase used the dative case,
maked in .ra. Occasionally, however, the locative-terminative case, marked in .e, is used
instead. This refects a further spread of the usage of the locative-terminative; such a use of
the locative-terminative for the dative does not appear in earlier Sumerian.
The combination of the Ikl of the genitive marker with the lel of the locative
terminative is written with the ke4 -sign, just as is the combination of the Ikl of the genitive
marker with the I el of the ergative. (This ambiguity of the ke4 -sign can lead to confusion.)
4. Here and in line 1 1 , the rot for "to live" is written !-g, not tU. That is, the root is
written syllabically, and not logographically. The spelling nam-!-g also occurs in the
Gudea inscriptions. Such spellings show us that the fnal 11/ of the root was indeed
pronounced in word-fnal position.
Lines 4ff. are a complicated genitive:
nam.til.Ulgi
J
.ak.e
nitabkalaga
lugal. Urim.ak
9. Because the locative-terminative is used in line 3 in the sense ofthe dative, it is resumed
by na, the dimensional-prefx properly belonging to the dative. In its more usual uses, the
locative-terminative is either not resumed at all, or else it is resumed by ni - properly the
dimensional-prefx belonging to the locative.
There is no expressed direct object (patient); the votive cylinder-seal itself is the direct
object.
The last word in the line is an anticipatory genitive, beginning a new sentence: "of this
cylinder-seal, its name is . . . ", that is, "the name of this cylinder-seal is ... ". This use of the
anticipatory genitive is similar to that seen in Text 1 5: alam-ba . . . mu-bi-im, "the name of
this statue is . . . ".
Line 9 includes the fnal word of one sentence, and the beginning word of the
following sentence. It is unusual in Sumerian orthography for one line to contain elements
of two different sentences. However, the space constraints in cylinder-seals occasionally
cause odd placement of signs within a line or case.
10- 1 1 . These lines express the actual name of the votive cylinder seal.
10. lugal-m
l
O
is probably a vocative. As in Text 21 , the vocative is unmarked.
The use of nam to form abstracts has occured several times. However, some
abstracts are formed with ni g, instead of nam: ni g-Sag
5
'
"goodness, favor". The original
meaning of nig is "thing", "something". It is relatively uncommon for nig to form such
abstract nouns. More commonly, it forms concrete nouns from verbal roots: g!7
'
"to eat";
ni g-g!7
'
"food". Therefore, it might perhaps be best to regard the ni g-Sag
5
in Text 22 as
concrete, rather than as abstact. Sollberger, in fact, tanslates this expression (in another
text) as: "grace, favour (in a concrete sense, 'good things
'
)" ( 1 966: 1 58).
Unlike nam, nig is not used with nominal roots. That is, such forms as *nig-l! gal do
not occur.
The transliteration ne assumes the interpretation as .ani.e, the possessive-suffx
followed by the locative-terminative case-marker, here with its more original meaning of
"at, by, through". A similar case occurred in Text 20, where the writing ne represented the
252
r.--::
possessive-suffx followed by the ergative case-marker .e. As discussed there, some
Sumero10gists would prefer to tansliterate such phrases by ni: getug
3
-nig-ag
5
-g-ka-ni.
1 1 . g is the modal-prefx of the cohortative mood. This is used for positive wish for frst
person, both singular and plural: "let me/us, may I1we". It is thus the frst-person
equivalent of the third-person desiderative modal-prefx b.
The view of Sumerian grammar presented in this book assumes the presence here of a
conjugation-prefx 1. which has assimilated into the modal-prefx g. However, since a
writing such as *g-l seems never to occur, the assumption of such assimilation may be
simplistic. Therefore, the idea that a conjugation-prefx must be present in every fnite
verbal form may need modifying.
One of the thorier problems in Sumerian morphology is the form of the root
appearing after the different modal-prefxes. In the singular, the cohortative g is always
used with the bamn!-root, whether the root is being used tansitively or intansitively. (The
plural is unsure.) When the cohortative is used in the singular, there is no overt marking
for person; that is, the subject (or agent) is not marked. But the direct object (patient),
however, is expressed by a personal-affx immediately preceding the verbal root. Thus, the
.n before the verbal root here marks the patient, and hence the root t-il must be interpreted
as transitive: "let me keep him alive/well".
The syntax of the personal-affxes of the cohortative (and of other moods) differs from
that of the indicative. In the indicative, the pre-verba1-root slot marks the agent in the
bam
n
!, but the patient in the mam. But in the cohortative, the agent is unmarked; the pre
verbal-root slot marks the patient, and the root is always in the bamty.
12. The enclitic copula is not used, although it was so used in the parallel expression in
Text 9: a1am-ba . . . mu-bi-im.
Discussion: sign formation
I Lesson 9, the use of gunu-strokes to form new signs was discussed. A similar
device was the addition of eig-stokes to a sign. Whereas gunu-strokes consist of short
lines, eig-stokes look more like a kind of cross-hatching. For example, the da-sign is
originally a picture of a hand-upper-shou1der-arm: _. The word has such meanings as
"forearm, side", etc. To indicate ., meaning specifcally "arm" (and then, by extension,
"stength"), eig-stokes appear on the part of the da-sign which approximately repre
sents the arm, yielding: .
The etymology of eig is unknown, but it must be connected with the word e, "bar
ley", whose cuneiform sign resembles these cross-hatchings.
-Syllabic writings
Syllabic writings of words usually written 10gographically (such as nam-t-
i
for nam
til) are not common in the Ur III royal inscriptions; it is not sure why they occur. By way
of contast, they are not uncommon in the two very large Gudea inscriptions.
Falkenstein believes that the relatively high frequency of such spellings in these two
Gudea texts shows that they were written down on the basis of a dictated text. That is, the
scribe(s) did not work from a written, already-prepared source, but rather the scribe(s)
Lesson ::
2
53
listened to the text being read, and wrote down the text as they went along. This idea does
seem like the best explanation for cerain kinds of eror which occur in the Gudea texts.
There is also some other evidence to suppor this view.
-Textual interpretation
The interpretation given above of the verb form in line 1 1 rests on the assumption that
the .n in the prefx chain refers to the patient. This view would probably be accepted by
most Sumerologists. However, other Sumerologists are less categorical in their thinking,
and would say that at our present state of knowledge, other possibilities cannot be
excluded. Therefore, this particular line has been tanslated as: "let me live by his ear of
favor", or even "let me make well his ear of favor". These tanslations refect different
possibilities of understanding the .n: marker of the frst-person intansitive verb, or marker
of the frst-person tansitive verb, or marker of the third-person patient, or reduced form of
the dimensional-prefx cross-referencing the locative.
-g and gae
The cohorative modal-prefx is /ga/, and the frst person independent pronoun is
/gae/. Since one begins with /g/ and the other with /g/, it does not seem that they are
directly related to each other. On the other hand, the similarity in form, and the fact that
both are used for functions involving the frst person, makes one pause. It is less easy,
however, to see such surface-level etymologies with the other modal-prefxes.
-Noun formation
The term zabar-dab
5
0riginally meant something like "one who holds the bronze". The
word for "scribe", dub-sar, originally meant "one who writes a tablet". In Text 22, g!-za
hi, "chair-bearer", occurs. A number of names of offcials in Sumerian are composed of an
active participle and an incorporated direct object. Many were borowed into Akkadian as
simple nouns: zabardabbu, tuparru, guzalU, etc.
- Seals
It is usually assumed that Kilula was the person who had this cylinder-seal fashioned.
He dedicated it to Meslamtaea, to bring life to Shulgi. Presumably Kilula gave it to Shulgi,
who then gave it to the temple. Thus, the cylinder-seal was designed to curry favor with
the king.
It is hard to say why the particular god Meslamtaea was invoked on this seal. Hallo
points out that "in private ex-votos inscribed on behalf of the king, it is not always cerain
whether the deity involved is the personal god of the king or his donor" ( 1 966: 1 37 n.53).
Without more precise knowledge of how such dealings took place, it is diffcult to say
exactly what the name of the seal does mean. l!gal, to illustrate part of the problem, can
mean "king, lord", referring to the god Meslamtaea. It can mean "king", referring to
Shulgi. It can even mean "owner", referring to the owner (which one?) of the cylinder-seal.
And, the loative-terminative (presumed! ) at the end of the nominal phrase can have several
different values: "at, by, through", etc. A somewhat materialistic interpretation of the name
254 Lesson ::
would be: "let me make him well at his ear of favor", that is, "at his ear which hears and
grants favors". The sense is, "let him listen favorably to me", "make him accessible to me".
However, one can
'
t help wondering whether Shulgi had an ear-ache.
-Votve seals
Votive-seals were basically non-functional, in the sense that they were not primarily
designed to be actually impressed upon written documents, as were the functional seals.
Rather, they were votve objects, in the form of a seal. There are very few examples of
votive seals actually being used (although it is always possible that this is an accident of
discovery).
Gelb says:
The main characteristic of the votive seals is that while they identifed the
donor of the seal, they were not used by the donor but by the divinity to
whom they were offered. Certain seals can b used for purely oramental
purposes but nothing would prevent the temple fom employing them in
identifying and legal purposes ( 1 977: 1 1 2).
Votive seals are also generally larger than functional seals. And, whereas functional
seals are inscribed in reverse - so that the impression comes out correctly - votive seals are
not; they are meant to be looked at, not to be used.
LmSn2
2JJ
Text 22a
supplementa
256 Lesson ::
Sign-list and vocabulary
Ia-a-ba-me-er PN. The etymology is unsure, but it is probably Akkadian. Landsber
ger has interpreted it as /baib-amir/, "he was wanted (and then) was seen", two verbal ad
jectives in the predicative state.
U-ku-un-
d
Zuen The name is Akkadian, "Sin has placed/established", Bkun Sin. Its lo
cation is unknown; it may have been close to Ishchali.
Commentary
6. The ki-sign is not in its expected place. Presumably, the scribe wanted to keep the en
sign and the zu-sign together. This particular writing may be regarded as a refection of the
practice of earlier periods in Sumerian, when the order of signs within a line or case was
not as fxed nor as linear as in later times.
Discussion
According to Hallo, "This is presumably the frst Ur III text ever published. It was
first copied by Charles Bellino in 1 820".
L0880u 23
This lesson reviews and summarizes certain main points of Sumerian grammar, so that
its broad stucture will not have become obscured by all the details presented in the
previous lessons. Secondly, specifc areas of disagreement among Sumerologists will be
pointed out.
A. General stucture
1 . Word-order
Sumerian is basically S-O-V in word order:
Nl -case marker N2-case marker
( 1 ) (2)
N3-case marker VP
(3) (4)
The N marking the agent usually comes frst. The constituents after the agentive NP
are more free; adverbial cases may precede the patient, or the patient may precede the
adverbial cases. Deviations from this standard syntax (for example, a benefactive at the
beginning of a sentence) are for emphasis.
2. NP structure
An NP can be of any length, and can be composed of many different structures: noun;
noun-noun compound; noun+adjective; genitive phrase; noun+possessive-sufx; etc.
These can be combined in different ways. A typical example of a complicated nominal
phrase is: e-gal-zu-!-ne, for e.gal. zu.ene, "your older brothers".
3. VP stucture
The basic structure (omitting some rare optional elements) is:
modal-prefx
( 1 )
conjugation-prefx
(2)
ROOT
(5)
personal-afx
(6)
4. Ergativity
dimensional-prefxes
(3)
personal-afx
(4)
It is difcult to see all the manifestations of ergativity in the Ur III royal inscriptions,
because very few intansitive verb forms occur, especially in the maru. Here ergativity will
be summarized, even if some of the constructons do not occur in this corpus.
The following four sentences will illustate: ( 1 ) the bamm-transitive; (2) the maru
transi ti ve; (3) the bam! -intansi tive/passi ve; (4) the marf-intransi tivel passi ve.
bam!-tansitive:
( 1 ) The king built the house.
luga1.e e.0 mu.n.du.0
257
258 Lesson :
manl-transitive:
(2) The king will build the house.
lugal.e e.0 Lb. du. e. 0
bam!-intransitive/passive:
(3) The king went.
lugal. 0 mu.gin. 0
manl-intransitive/passive:
(4) The king will go.
luga1 . 0 Ldu. 0
In both the bamm and the manl, in both the transitive and the intransitive/passive, the
nominal participants are marked in the same manner: the agent is marked in .e (the ergative
case-marker) and the patient is marked in . 0 (the absolute case-marker).
Thus, looking at the manner in which the nominal participants in the sentence are
marked, Sumerian is an ergative language: the agent is marked in one manner (. e) , and the
patient in another (. 0).
In terms of cross-referencing by the personal-affxes, however, the situation is
different. With a transitive verb, in the bamm the agent is cross-referenced by the personal
affx slot before the root. In the maru the agent is cross-referenced by the personal-affx
slot after the root. In the bam.! the patient is cross-referenced by the personal-affx slot
after the root. In the maru the patient is cross-referenced by the personal-affx slot before
the root.
With an intransitive/passive verb, the bamm and the maru behave the same way: the
patient is cross-referenced by the personal-affx slot after the root.
This means that the cross-referencing system used in the maru behaves differently in
the transitive than it does in the intransitive/passive. Repeating the previous sentences in d
differen t order,
(5) The king built the house.
lugal.e e.0 mu. n. du. 0
(6) The king went.
luga1.0 mu.gin. 0
(7) The king will build the house.
lugal.e e.0 Lb.du.e.0
(8) The king will go.
lugal. 0 Ldu. 0
In the bamt.!, the patients in (5) and (6) are cross-referenced in the same manner: by
the personal-affx .0 after the root. But in the maru, the patient is cross-referenced in two
Lesson 23
259
different ways: In (7), the patient is cross-referenced by the personal-affx slot before the
root. In (8), the patient is cross-referenced by the personal-affx slot after the root.
Thus, the patient in (7) (the direct object of the tansitive verb) and the patient in (8)
(the subject ofthe intransitive/passive verb) are not teated the same. Therefore, the system
cannot be considered ergative, since the defnition of ergativity is that these two patients be
treated the same.
This argument hinges on the defnition of ergativity. In the bamn! and the maru, the
nominal participants are marked in an ergatve manner. It is only in the system of cross
referencing that the two patients in () and (8) are teated differently from each other. If
one accepts the idea that ergativity does not just refer to the nominal markers, then we are
forced to say that the maru does not behave in an ergative manner. But since the bamtg
does behave in an ergative manner, Sumerian must be called a split ergative language, split
along d aspectual axis. However, if we look only at the nominal markers and not at the
system of cross-referencing, we can say that Sumerian is ergative, and not have to refne
this term (at least, in terms of an aspectual axis; there is some indication that Sumerian is
split along a nominal - pronominal axis).
(5) Personal-affxes
The interpretation of the personal-affxes presented here has been basically known
since Poebel. This interpretation was made much more explicit, and placed in an ergative
framework, by Michalowski ( 1 982). While this interpretation seems to work for most
verbs, it is clear that it does not work for all of them. For example, a commonly occurring
sentence is: kiib PN.ak ib-ra, "the seal of PN was rolled". The verb apparently is intran
sitive/passive, as shown by the lack of an ergative case-marker .e, yet the . b in the pre-root
slot seems to cross-reference it. The proper analysis of such forms is still unclear to us; it
may be much more complicated than it frst appears. More work remains to be done on
explaining and categorizing the exceptions which seem to occur with the personal-affxes.
(6) Root
Sumerian roots can be divided into two classes: nominal roots, such as lu, and verbal
roots, such as sar; there is no morphologically distinct class of adjectival roots. Adjectives
are to be regarded as participles (or something similar) of verbal roots. For example, gibil
can be used as an adjective meaning "new", from the verbal root meaning "to be new".
There is no canonical shape of the verbal root. Most of the verbal roots occurring in
the texts in this book are of the type CV (!, g4), or of the type eve (g!, pad). How
ever, there has been one eveve root (Wig), one of the type vev C!), one unsure (ba
il, and one borrowed from Akkadian (gi-i). Similarly, there is a fairly wide variation in
the shape of the nominal root. They have taken the for V @; CV Cl ); VC (ur); eve
(g
g
); vev (utu); evev (dumu); veve (alam); eveve (temen); etc.
The root is unmarked for such categories as active - passive, tansitive - intransitive,
causative - passive; etc. For example, ti l can mean "to live" or "to let live"; kur
9
can mean
"to enter" or "to make enter", "to bring in". In Text 1 5, gub is used in the meaning "to
stand"; in Text 6, it is used in the sense of "to plant" a garden.
260 Lesson 23
B. Aeas of disagreement
There are a number of disagreements and alterate explanations about certain features
of Sumerian grammar. Because a knowledge of these alterate views is presupposed by
Sumerologists, it is important to be at least familiar with the main differences fom the
views presented here.
1 . Verbal phrase
The interpretation of the Sumerian verbal phrase presented in this book owes much to
the ideas of Gene Gragg (briefy sketched in [ 1968]). Two principal differing views are
those of Falkenstein and Jacobsen. Yoshikawa
'
s views also differ considerably, but he has
not yet published a complete synthesis of his views.
(a) Falkenstein
His views are sketched in ( 1959) , and particularly adumbrated for the Gudea texts in
( 1 978
2
). His interpretation of the prefx chain is as follows:
bi.
Prlformative Konjugationsprlfxe Prafe Verbalinfxe Wurzel
( 1 ) (2) (3) (4) (5)
Pr!rmative: Correspond to the modal-prefxes.
Konjugationsprafxe: Include only the following: l.mu, and al.
Prafxe: Correspond to the dimensional-prefx na, and the conjugaton-prefxes ba and
The conjugation-prefxes limmi/, limma/, etc. , are considered by Falkenstein to
represent *i-bi-i and *i-bi-a respectively; that is, they contain both a Konjugationsprafx and
a Prafx.
Verbalinfxe: Include the dimensional-prefxes except na, and the personal-affxes.
Falkenstein
'
s views of Sumerian grammar have been criticized on two basic grounds:
a mixing of synchronic and diachronic description, especially in the categories of Kon
jugationsprafxe and Prafxe, and a regular use of terms and concepts proper to the Indo
European and Semitic languages, but which do not necessarily apply to Sumerian, or to
languages which are typologically similar to it.
(b) Jacobsen
His views are spelled out most explicitly in ( 1 965); some important comments occur in
his review of Thomsen ( 1988). His terminology for the prefx chain is based on formal
criteria:
Profxes
( 1 )
Prefxes
(2)
Infxes
(3)
Root
(4)
Lesson 23
Profxes: Correspond to the modal-prefxes.
261
Prefxes: Correspond to the conjugation-prefxes. However, Jacobsen has defnite
and specifc views on the meaning and function of the conjugation-prefxes. He believes
that they can be analyzed into two morphemes: "a pronominal-adverbial element denoting a
region (m, b) and a relater or case-mark (u for tangentive, a for illative, i/a for allative)."
This analysis is very detailed.
Infxes: Correspond to the dimensional-prefxes and personal-affxes, in a rather
complicated way.
In general, Jacobsen's analysis tends to be much fner than that of other Sumerologists.
In some cases, vocalic (or consonantal) alternations which most Sumerologists would
regard as phonetic or morphophonemic alteration are regarded by Jacobsen as repre
senting different morphemes. He has been criticized for producing much too detailed a
segmentation of the morphology. On the other hand, Jacobsen's interpretations are based
on a close empathy for the texts; perhaps more so than any other living Sumerologist,
Jacobsen has a feeling for and understanding of the content of Sumerian texts, especially
literary texts.
(c) Yoshikawa
Yoshikawa has published a number of articles in an attempt to defne the functions of
the conjugation-prefxes. In some of the later articles, he has tried to use phenomena from
other ergative languages, especially Georgian, to elucidate Sumerian. In his 1 981 article, he
says:
We can tentatively specify the function of the respective Sumerian verbal
prefxes as follows:
1. bi -: Ientive locative/superessive prefx
im-mi- : Ventive locative/ superessive prefx
ba- : Ientive reflexive (/subjective) prefx
im-ma- : Ventive refexive (/subjective) prefx
II. mu-: Topical agentive (/objective) prefx
i - : Non-topical agentive (lobjective) prefx
III. al- : Neutral prefx.
2. Maru-infection
maru-forms such as l-sar-re have been analyzed here as Lsar.e.0. The root i s sar,
which is a member of the affxation-class of maru-formation; .e is the maru-suffx; .0 is
the personal-affx cross-referencing the agent, that is, the third person marker.
Edzard believes that l-sar-re is to be analyzed as Lsar.e. sar is a member of the in
variable class of maru-formation, and . e is the third person marker.
The frst analysis sees two morphemes after the root; the second sees one. It is not
easy to resolve this issue. The problem is partially the result of a lack of evidence. Far
fewer intransitive/passive verbal forms are preserved than transitive forms. Fa fewer
present-future forls are preserved than past. And, fewer frst and second persons are
262 Lesson :
preserved than third, and fewer plural forms than singulars.
Many of the examples preserved which show the greatest amount of grammatical
variation are attested in relatively late copies of literary texts. In these texts one must
always guard against misunderstandings by the Akkadian-speaking scribes who copied
down these texts. Similarly, the Akkadian-speaking scribes who drew up the grammatical
texts discussed in Appendix 2 sometimes include tansitional or analogical forms, which
cannot be regarded as conforming to the standards of earlier Sumerian.
The problem is also caused by ambiguities in the writing system. Types of ambi
guities include: ( 1 ) The writing system cannot distinguish between a form such as i . sar.en
or Lsar.e.en; they would both be written as l-sar-re-en. Isolated writings of the type i-sar
re--en do not resolve such ambiguities, because there are several possible explanations for
the full writing. (2) /d! is an amissable consonant. Therefore, a writing such as i-sar-re
can stand for i . sa.e, i . sar.ed, or Lsar.e.ed. (3) The de-sign i s ambiguous; it can stand for
de or for ne. This means that a writing such as i-sar-re--de can represent several different
possibilities: Lsar.ed. e. , Lsar.e.ed, Lsar.ene, Lsar.e.ene, etc. It is ambiguities such as these
which make it difficult to correctly analyze the morphology of the maru.
3. Manl classes
The number of different types of formation of the maru from the bamt! is unsure.
Yoshikawa's initial formulation showed three classes: ( 1 ) affixation (2) reduplication (3)
alternation. This scheme works well for the Ur In royal inscriptions, but this is partially
because only a very limited number of man forms occur in these texts. Y oshikawa himself
has said that his classification system will need expansion and revision.
Edzard modifed and extended the system into fve classes: ( 1 ) unchanging. Since
Edzard considers .e to be a third person subject marker, not a marker of the marG, this is
actually the same as Yoshikawa's affxation group. (2) reduplication (3) root-varying.
This is a sub-class of Yoshikawa's alternation class; the two roots are different, but phone
tically similar in some way. (4) alternation (5) iregular. These last do not seem to fi t into
the fst four classes.
As more progress is made in Sumerology, the last class will be further refned. Some
of the verbs which now seem irregular will eventually be shown to follow rules which are
not yet known. Whenever linguists study an unknown language, they are apt to see more
irregular forms than when they have been able to examine the data more thoroughly.
4. Normalform
There has been some discussion about the conjugation of the intransitive/passive verb
in Sumerian. Poebel believed that there was one conjugation for the bamt! intransi
tive/passive, and a different conjugation for the marG intransitive/passive. Most modern
Sumerologists, however, believe that there is only one conjugation for both the bam
n!
and
the manl (that is, there i s only one set of endings), although the particular root used (i.e.,
either bam
n!
or marG) will differ. This one conjugation is usually called the "Normalform"
(even though this is a rather meaningless term). What follows is Poebel's reconstruction of
the bam
n!
intransitive/passive and the marG intransitive/passive.
as:
Lesson 23 263
For Classical Sumerian, Poebel reconstructs the intransitive/passive form of the bamt.!
frst person singular
second
third
frst plural
second
third
i-sar-en
i-sar-en
i-sar
i -sar-re-en-de-en
i-sar-re-en-ze-en
i-sar-re-eS
He reconstructs the intransitive/passive of the maru as:
frst person singular
second
third
frst plural
second
third
i-sar-re-de-en
i-sar-re-de-en
i-sar-re
i -sar-re-de-en-de-en
i-sar-re-de-en-ze-en
i -sar-re-de-eS
i . sar.en
Lsar.en
i . sar.0
i . sar.enden
Lsar.enzen
Lsar.eS
Lsar.ed.en
Lsar.ed.en
Lsar.ed.0
i . sar.ed.enden
i . sar.ed.enzen
Lsar.ed.eS
That is, the man} intransitive/passive differs from the bamm intransitive/passive sole
ly in the presence of the element .ed. Poebel believed, however, that this sytem worked
only for Classical Sumerian. Under the infuence of Akkadian, several analogical changes
took place in the system of the man} intransitive/passive, eventually yielding a completely
new paradigm:
frst person singular i-sar-re-en i . sar.en
second i-sar-re-en i . sar.en
third i-sar-re Lsar.e
frst person plural i -sar-re-en-de-en i . sar.enden
second i -sar-re-en-ze-en Lsar.enzen
third i-sar-re-ne i . sar.ene
The .ed element has disappeared; the third person singular is now marked by .e; and
the third person plural is now marked by .ene. Thus, the endings have been assimilated to
those of the maru transitive.
From the period between the time of the paradigm of the more classical period and the
time of the newer paradigm, several transitional forms are attested (for example, i -sar-re-de,
Lsar.ed.e) for the maru intransitive.
PART TREE: APPENICES
PQQ0u0X 1
Hstor
The perioization of Sumerian, like the perioization of Akkadian, is to some extent
based on non-linguistic criteria, such as political and historical events. The essence of this
periodization is:
I. Archaic Texts (31 00 BC)
I. Archaic Sumerian (3100-2600 BC)
II. Classical Sumerian (2600-2300 BC)
IV. Neo-Sumerian (2300-200 BC)
V. Post-Sumerian (200 BC- l O AD)
I. Archaic Texts (31 00 BC)
The earliest known tablets containing writing come mostly from Uruk (whose modem
name is Warka). They come from a stratum usually designated as Uruk IVa, commonly
dated to about 31 00 BC. They were frst assigned to "Uruk IVb", but then redated to
"Uruk IVa" (later than Uruk IVb). Secondary literature often confuses this, and so some
accounts speak of Uruk IV, some of Uruk IVa, and some of Uruk IVb.
Almost all of the tablets were found in a "dump" inside the Eanna temple complex.
This means that it is very difcult to date the tablets, either in terms of relative chronology,
or in terms of absolute chronology. According to Eva Storenger, a further complication
is the fact that to some degree, the architectural levels and building phases at Uruk have
been dated on the basis of the tablets found, not vice-versa. (This is also tue of some
tablets found later than Uruk IV.) She says that
The phases of writing are everywhere used in order to date the level in which
the specifc tablets were found. This leads to the conclusion that the
possibility of the existence of tablets prior to IVa . . . has not been recognized,
but that deposits in which a few tablets were found have been dated to IVa
on that basis alone ( 1980:48 1).
Other scholars, however, are less willing to agree with her implicit criticism of the
reconstruction of the archaeological history of the site.
The frst of these texts were excavated by the German archaeologists of Uruk during
1 928- 1 931 . In 1 936, Falkenstein published his Archaische Texte aus Uruk, a seminal
work. In it he treated the frst 620 tablets found. By now, upwards of four thousand of
these texts are known, found during subsequent excavations. The majority of these (mostly
fragmentary) tablets have not yet been published.
There are a few similar tablets from other sites, mostly in northern Syria. In addition,
both from Uruk and especially from Syra (and elsewhere, including Iran) there are a
number of what are commonly called "numerical" tablets. It was at frst thought that these
265
266 Appendix 1
tablets represented numbers, but more likely they indicate items being counted; their precise
intepretation is unsure. Two fom Tell Brak in norther Syria, found in 1984, in particular
are quite archaic looking. Based on achaeological criteria, these numerical tablets do not
seem to be any older than the pictographic tablets discussed above; rather, both numerical
and pictographic tablets ocur in U ruk IV a.
A very achaic-loking tablet comes from Kish. Unfortunately, it is from an
uncontolled context, and so it cannot be dated achaeologically. This tablet is frequently
referred to in popula literature about the Ancient Near East as being one of the very ealiest
tablets known (if not the ealiest tablet), but it has a fairly elaborate division into cases,
which makes it more likely that it is later than Uruk IV.
These early texts ae undeciphered, and perhaps to some degree undecipherable.
Therefore, it is impossible to be certain about what language they are written in. There are
several reasons why these texts cannot yet be read:
- Even at this ealy date, the supposedly pictographic nature of the signs is not always
obvious. Most of the signs ae already abstact; it cannot be determined what they were
originally meant to depict.
- Some of the abstact signs can be understoo on the basis of knowledge of later
Sumerian. However, a fair number of the signs (perhaps 30%-50%) cannot be read or
understood. These ae signs which eventually passed out of use, so that there is no later
grammatical tadition to provide information about their meaning.
- Most of the signs which can be understood d logographic; in theory, these can be
read in any language. A sign which is a picture of a mountain, for instance, could be read
as "mountain", "Berg", adf, kur, etc.
- There do not appea to be any syllabic signs. This means that no grammatical features
can be seen; for example, there ae not any case-makings on nouns. There do not appea
to be any verbs at all. Writing at this stage was a highly mnemonic device.
- These ae mostly administatve records, sometimes very short - occasionally, just a
few signs long. It is very diffcult to understand such texts out of context, that is, without
knowledge of the administative framework which produced these texts.
In spite of these problems, most scholas think that these texts ae written in Sumerian.
The main reason is because texts have been preserved from the later Uruk III statum,
which is known to be Sumerian. Since archaeologists see a cultural continuum between
Uruk IV and Uruk Ill, it is reasonable to assume that the same language is present in both
stata. In addition, Powell has agued that the system of metology used in these early texts
seems to be the same system used in clealy Sumerian texts:
The system of numeration deducible from the notation present on Uruk
IVa/III tablets makes it virtually certain that these tablets are written in
Sumerian and, ipso facto, highly prbable that the inventor of the pictorial
writing system was also a Sumerian ( 1981 :423).
These tablets are being studied by Magaet Green and Hans Nissen (a student of
Falkenstein), and ae in the course of publication (preliminary discussions Nissen 1985,
1986; frst major publication Green and Nissen 1987). Although most ealier scholas
differentiate between Uruk IVa and Uruk Ill, Nissen subsumes both into one category,
Histor 267
"Archaic Texts". (The fgure of "4000 texts" cited above therefore includes tablets from
both Uruk IVa and Uruk ll.) Nissen estimates that about 85% of the texts are economic
records, and about 1 5% are lexical lists. He is more optmistic than most scholars about the
possible decipherment of these texts, blieving that he can identify about 700 of 1000
different signs, and that the texts are "possibly" written in Sumerian. He bases his
arguments partially on the continuity btween the early lexical lists (unknown at the time of
Falkenstein
'
s publication) and later, well-understo lexical texts.
Nissen thinks that the texts from Uruk can b divided roughly into two classes, one
representing an "early stage of the script", and one a "younger stage of the script" (these
two divisions do not exactly correspond to the taditional Uruk IVaUruk II divisions).
The latter, consistng of most of the tablets found since Falkenstein
'
s publication, are more
amenable to analysis.
The fact that many of the signs in these early tablets are already abstact has led many
scholars to assume that there was some previous development bhind the signs. That is,
these tablets do not represent mankind
'
s frst attempt at writing. Several different
hypotheses have been proposed:
- The Sumerians may have borrowed their writng system fom some other people,
perhaps some distance away from Mesopotamia. This is not impossible. It has often been
argued that the Sumerian writing system does not ft the Sumerian phonological system
very well; this might imply that the writing system was created for a different language.
This particular theory has been around for many years; it is obviously very difcult to
prove.
- Earlier writing may have been on perishable material, such as wood, or animal skins,
or palm leaves, etc. There are parallels to such practice fom later Mesopotamia, and from
Arabia around the time of Muhammad. This is also a rather old theory, but it is also
virtually impossible to prove.
-
I a series of articles beginning in 1 977, Denise Schmandt-Besserat has argued that
the earliest "precursor" of writing was clay "tokens", which have ben found at various
sites throughout the Ancient Near East, starting from the early Neolithic. Writing
originated in a conceptual leap, from the use of physical tokens, to the use of symbols to
represent these tokens: "The substitution of signs for tokens was no less than the invention
of writing" ( 1 986: 37). She envisages the following stages ( 1 986: 35):
( 1 ) 8000 BC appearance oftokens
(2) 3250 BC clay envelopes hold tokens of particular tansactions
(3) 3200 BC signs are impressed on the surface of envelopes
(4) 3 100 BC clay tablets appear with impressed and incised signs
Powell agrees:
Cuneiform was invented in a short period of time around 3000BC by a
citizen of the Sumerian city of Uruk . . . . It arises conceptually out of the token
system described by D. Schmandt-Besserat. ... The pictorial ancestor to
cuneifor writing was invented as a conceptual whole during the tme period
represented by the Uruk I -Ill archaeological stata ( 1 98 1 :419-420).
268 Appendix 1
Contast this with the more "evolutionary" thinking of Walker:
Thus it is bginning to lok as if we should think in ters of the invention
of writing as bing a gradual proess, accomplished over a wide area, rather
than the product of a single Sumerian genius ( 1986:9).
Nissen hints at the complexity of the developing of writng: "Writing was developed
at the end of the fourth millennium B.c. by a mixed language group in which Sumerian
was apparently the main component" ( 1 988: 14).
The most vocal critic of Schmandt-Besserat
'
s view has ben Lieberan ( 1 978, 1980).
However, probably the majority of Sumerologists agree with her overall interpretation of
the development of writing from tokens.
11. Archaic Sumerian ( 3100-2600 BC)
Tablets from this stage are found at several sites. The oldest are fom Jemdet Nasr;
others are from Uruk Ill, Uruk 11, Ur, Fara (ancient Shuruppak) , and Tell Abu Salabikh
(2600 BC). These dates are not exact, and, as discussed above, Nissen includes both Uruk
IVaJUruk III together; the ramifcations of this revised relative chronology are still to be
worked out.
I the Jemdet Nasr texts, there is a personal name written:
d
En-lil-1. The word for
"arrow" in Sumerian was ftil; the sign used to represent this word was originally a picture
of an arrow. It is doubtful, however, if this name means something like "Enlil is an arrow".
But, the root ftil in Sumerian also means "to live". Thus, the name
d
En-lil-1 means "Enlil
lives", or more likely "May Enlil give life", "May Enlil keep alive". That is, in this name a
cuneiform sign is being used syllabically: The 1-sign is being used stictly for its phonetic
value, not for its logographic value. This writing shows that the script is being used for a
language where the words for "arrow" and "to live" are homophonous.
This interpretation of the personal name
d
En-lil-1 goes back to Falkenstein, in 1936.
It gained immediate acceptance by Assyriologists and Sumerologists, who believe that the
writing: ( 1 ) shows the existence of phoneticization; (2) shows the personal name to be
Sumerian; and (3) shows the language of the texts in which the personal name occurs to be
Sumerian. (The same name may also occur in the later phase of Nissen
'
s "Archaic Texts".)
In 1974, however, A.A. Vaiman suggested that the name should be read as E.EN.TI; he is
followed by Lieberan.
It is possible that there are other instances of phoneticization in the texts from Jemdet
Nasr, but the evidence is not unequivocal. Texts from the later stages of this period show
increasing phoneticization, and are clearly Sumerian.
Up until the 1960s, virtually all of the texts which were known fom this period were
the usual administative and economic texts. These are not always easy to understand,
again because of the lack of any context. This situation was changed in 1963. In that year,
the University of Chicago began excavations at a site called Tell Abu Salabikh, near
Nippur. It turried out that the majority of tablets and fagments found were literary texts.
Some were compositions which were known fom later times. For example, there is a text
known as the "Kesh Temple Hymn", preserved in numerus Old Babylonian copies dating
to about 1 800 BC. Fragments turned up at Tell Abu Salabikh - some eight hundred years
History 269
earlier. Other texts turned out to be previously unknown compositions. For example, one
is a collection of temple hymns. Most of these texts are scarcely intelligible; not much is
known about literary Sumerian of this period.
Thus, the primary importance of Tell Abu Salabikh lies in the existence ofliterary texts
from the middle of the third millennium BC. Since these discoveries, scholars have
recognized fragments of literary texts among some tablets which have been known for
many years. For instance, some of the Fara texts are fagments of proverbs which are
known from later proverb collections.
The texts from Tell Abu Salabikh are also important, because a number of the literary
texts (and lexical texts) have colophons of the sort: "so-and-so wrote". It i s not known
what the word "wrote" means here exactly: Does it mean that the scribe "composed" the
composition, or that the scribe "copied" the text from a master tablet, etc. However, what is
interesting is that a number of these scribes have demonstrably Semitic personal names.
It is diffcult to date the intrusion of Semitic-speaking peoples into Mesopotamia, on
linguistic or other grounds. The frst evidence is usually thought to be the presence of
Akkadian loan words in early Sumerian. These loan words are diffcult to evaluate, how
ever, because it is not always certain which way the borrowing went, or whether a third
language may have mediated a word, etc.
The Semitic names in these colophons are thought to be the frst real evidence of
Semitic-speakers in Mesopotamia. If the Fara texts and the Tell Abu Salabikh texts are
dated to about 2600 BC, that gives a terminus ante quem for the arrival of Semitic speakers,
but it does not say anything about how long they might have been present in Mesopotamia.
If they had already become scribes, they must have been there for some time, since they had
worked themselves into the intellectual life of the community.
Ill. Classical (or Old) Sumerian (2600-2300 BC)
Most of the texts of this stage come from Lagash, from a period known as the "First
Dynasty of Lagash". Besides the usual administative, economic, and legal texts, there are
a fair number of royal and private inscriptions. There are also some letters, and even a few
literary fragments are now known. Royal inscriptions are also known from other sites.
The end of this period corresponds to the rise to power of the Semitic-speaking
Dynasty of Akkad (2334-21 54 BC). As mentioned above, Semitic-speaking peoples must
have been present in Mesopotamia for centuries before the time of Sargon, the founder of
the dynasty (ruled 2334-2278 BC). It must be presumed that Mesopotamia was bilingual
during this time, at least to some degree. However, with a Semitic-speaking dynasty in
power, Sumerian gradually started to move into second place.
A recent addition to the corpus of texts known from this period are the texts from Ebla
(in northern Syria, therefore from outside of the Sumerian-speaking heartland). To date,
upwards of ten thousand texts and fragments have been discovered. The texts are in both
Sumerian and Eblaite; until the material is better studied, it is not sure which language
predominates. Early accounts said that perhaps ninety percent of the texts were written in
Sumerian; this is probably much too high a fgure. The problem is that the texts written in
Eblaite are couched in a Sumerian orthography, utilizing a large number of Sumerian
270 Appendix 1
logograms.
Most of the texts found at Ebla are administrative or economic, chiefy concerned with
the metal and textile industries. However, there are also lexical lists; some are Sumerian
lists known from later periods, others are bilingual Eblaite-Sumerian texts. There are a few
literary texts in Eblaite (mostly incantations); these are extemely diffcult to understand.
The existence of possible literary fragments in Sumerian is disputed.
It is still too early to assess the Sumerian texts from Ebla. It is clear, however, that
much new information is present. For example, the bilingual lexical texts include Sumerian
words and expressions not elsewhere attested.
IV. Neo (or New) Sumerian (2300-2000 BC)
Although Sumerian was on the defensive in the face of Akkadian, it enjoyed a strong
- albeit brief - revival under the kings of the Vr il Dynasty (21 1 2-2004 BC). This is the
period from which the most tablets of all have been preserved. There are texts from many
sites, including Vr itself, Drehem, Lagash, Larsa, Nippur, and Vmma. There are literally
thousands and thousands of mostly economic documents, as well as inscriptions, letters,
and other types of texts. Also, more and more tablets with literary texts are being dated to
this period.
From some time before the reign of Vr-Nammu (the founder of the Dynasty), there are
a fair number of inscriptions from the reign of Gudea, the local ruler of Lagash. Many of
these are inscribed on statues of Gudea himself. There are also two large cylinders of his,
inscribed with a very long building hymn. The largest ("Cylinder A") is almost one
thousand lines long; it is apparently the longest connected Sumerian inscription.
The dynasty of Gudea is referred to as "Lagash II". The chronology of Lagash II is
unsure; some see it as roughly contemporaneous with Vr Ill, but most view it as following
immediately upon the Old Akkadian period. In any case, the language of the Gudea texts is
more or less the same as that of the Vr III texts.
It is not known when Sumerian ceased to be a spoken language; this is a current topic
of discussion among Sumerologists and Assyriologists. Vsually assumed to be spoken
during the Vr III period, it was under the greatly increasing infuence of Akkadian. Some
scholars use the fgure 2000 BC, others 1 900 BC, for the date when Sumerian ceased to be
spoken, but this fgure is rather arbitrary. The language continued to be spoken by ever
smaller groups of speakers, and it is impossible to say when the last speaker of Sumerian
died. What is usually meant by the question "When did Sumerian die out?" is "When did
the native language of the people who produced the texts we have cease to be Sumerian?"
Pockets of native speakers of Sumerian may have continued for some time, but without
producing any texts.
Other scholars have argued for an earlier death. Jerrold Cooper has said that
"Sumerian as a spoken language was in all probability dead or nearly so in Ur Ill"
( 1 973: 241 ). His argument is based on the types of documents preserved during the Ur III
period. Both Kienast ( 1 98 1 a) and Michalowski ( 1987) essentially agree.
Jacobsen, on the other hand, says "We therefore assume that Sumerian was still
spoken as everyday language in the south in the Ur il period and a major part of the Isin-
Histor 271
Larsa period as well ( 1988: 124). Liebenan has stated that there is some evidence to show
that "Sumerian was spoken during the Old-Babylonian period" ( 1979:27).
V. Post-Sumerian (2000 BC- l O A)
This is occasionally divided into the following subivisions:
Early Old Babylonian
Later Old Babylonian
Post Old Babylonian
200- 1 80 BC
1 80- 160 BC
160-
After the Early Old Babylonian period, Sumerian was essentially dead as a living
language. However, it continued to b taught in the schools as a language of culture, and as
a language of religious importance. The parallel has frequently ben made with the role of
Latin in the Roman Catholic Church: Latin is still written, and even to some degree
spoken; hence, it i s "living", even though it is not spoken as a native language of anyone.
Recently, Vanstiphout has used the ten "Standard Sumerian" to mean
the language used in the literary documents of the Ur III and Old Babylonian
periods . . . . This language is a literary and therefore written fon, taught in
school for educational and literary purposes ( 1985: 1 ).
The majority of Sumerian literary tablets which have ben preserved are frm the Early
Old Babylonian period. However, there are also original texts written in Sumerian from
this period; examples include royal inscriptions (alongside those written in Akkadian), and
hymns written in honor of some of the Old Babylonian rulers.
Although the original composition of most Sumerian literary texts was in Sumerian,
by native speakers of Sumerian, the native language of the scribes who copied down the
literary texts during this period was Akkadian, not Sumerian. This led to a stong linguistic
infuence of Akkadian upon Sumerian, to the extent that the literary texts contain features
which would appear to be "wrong" by the rules of Classical Sumerian grammar.
Michalowski, for example, has spoken of the "profound changes in grammar evident in the
Old Babylonian literary texts" ( 1 980a:91); "during the Old Babylonian period
Mesopotamian scribes wrote Sumerian utilizing a profoundly different grammar, much
infuenced by Akkadian" ( 1980a:86 n.3). The extent of deviaton from the non vaies
frm one particular text to another. Inanna
'
s Descent, for example, is pretty good Su
merian, with only a few "wrong" verbal fonns. In Gilgamesh and Agg, on the other hand,
there are more "wrong" verbal fons than "right" ones.
It is, of course, always possible that fons which we regard as "wrong" are in fact
"right", but our understanding of Sumerian grammar is not yet sophisticated enough to
correctly interret such fons. J acobsen has emphasized this methodological point:
Once it has been decided that our sources are generally suspect it becomes
natural to see all unexpected and diffcult features as due to corrupton,
without seriously considering the possibility that our own limited and rough
knowledge might be at fault and need nision . ... The essential thing is to b
slow to dismiss diffculties with the easy assumption of mistakes by the
Ancients ( 1988: 1 25- 1 26).
272 Appendix 1
Sumerian continued to be written right down to the Christian era. These very late texts
are either cultic or astonomical. There are even a few Sumerian texts (including portions
of canonical lexical lists) written in Greek characters. The very latest cuneiform texts
preserved are several astonomical almanacs, written in Akkadian with a great number of
Sumerian logograms for technical ters. The latest of these can b dated by interal critera
to the year 385 of the Seleucid Era, corresponding to 74175 AD.
As mentioned above, this periodization (and most others) is to some degree based on
exteral (historcal and political) crteria, not on purely linguistic criteria. Jacobsen has
proposed a different scheme, based on linguistic criteria (without yet assigning precise
dates), while emphasizing the fact that the paucity of the data prevents overly-fne
subdivisions:
I Archaic
I OldSumeran
II Standard Sumerian (beginning with Narim-Sin of Akkad)
I LateSumerian
PQQ0B0X 2
Mesopotamia Souces
Much of our knowledge of Sumerian derives from the intellectual activity of the
Mesopotamian scribes themselves. This section describes some of these Mesopotamian
sources.
Lexical lists
It is especially in the area of lexicography that modem Sumerological studies depend
on native sources. From a very early period, the Sumerians began to compile "lexical lists".
These early texts were monolingual, consisting simply of lists, usually of words for
semantically related things: lists of names of fshes, of professions, of stones, etc.
Although most lexical lists are loosely arranged according to subject, others are organized
according to graphic shape, or even according to phonological shape. Fragments of such
lists occur among the very earliest Sumerian texts which have been preserved.
These early texts were the product of Sumerian scholars, originating in the Sumerian
scribal school system. Lexical lists become more and more common, however, beginning
about with the Old Babylonian period. At that time, Sumerian was in the process of
completely dying out as a spoken language - if it had not already done so. By the end of
the Old Babylonian period, if not earlier, Sumerian was only spoken in the schools. These
later lexical lists are a product of the Mesopotamian scribal schools; their purpose was to
aid the Akkadian-speaking scribes in their study of Sumerian.
By the late Old Babylonian period, many lexical lists assumed what is often called
"canonical" status; that is, they became standardized in content and in form. There are
about a dozen such "canonical series". Some are monolingual in Sumerian, like the earlier
texts, but most are bilingual; they have a Sumerian word in the left-hand column, and an
Akkadian equivalent in the right-hand column. Some have three columns: a phonetic
spelling of the sign; the S umerian logogram; and the Akkadian meaning.
Many of these series are quite extensive. One of the largest and best-preserved is
known (both to us and to the ancient Mesopotamian scribes) as "ur
S
-ra bubullu", after its
frst entry. In its canonical form, this series occupied 24 large tablets, totalling about
1 0,000 entries. Civil has called it an "inventory of material culture" ( 1 976: 1 25). He
descri bes its contents as:
(tablets 1 and 2): legal and administrative terminology.
(3-7): trees and wooden artifacts.
(8-9): reeds and reed artifacts.
( 1 0) : pottery.
( 1 1 ) : hides and copper.
( 1 2) : other metals.
( 1 3) : domestic animals.
( 1 4) : wild animals.
273
274
( 1 5) : parts of the body.
( 1 6) : stones.
(17): plants.
( 1 8) : birds and fsh.
( 1 9) : textiles.
(20-22): geographic tens.
(23-24): food and drinks.
Appendix :
The frst enty of this series has Ufs-ra (the Sumerian word for "interest-bearing loan")
in the lef-hand column, and the Akkadian gubullu (with the same meaning) in the right
hand column.
Lexical lists such as these help us to detenine the meaning of Sumerian words. Some
of the lexical lists go even further, and enable us to determine the reading (that is, the
approximate phonetic rendering) of a certain sign. For example, there is a relatvely late
lexical series known as "diri", which in its canonical fon occupied seven tablets, with
more than 2,000 enties. This series was used to give the pronunciation of compound
logograms, that is, logograms whose reading cannot be infered fom the individual pats
(such as zabar, written with the UD-KA-BAR signs; without lexical lists, it would be vir
tually impossible to deduce that the pronunciation of these three signs was Izabar/). For
this reason, such compound logograms d often refered to as "diri-compounds".
In diri, the pronunciation of the logogram under discussion is given in the far left -hand
column, using a resticted number of syllabic signs. Then comes the logogram in question.
Then comes the name of the sign (at least as early as the Old Babylonian period, the
Akkadian scribes gave names to the individual signs). Finally, the last column gives the
meaning of the sign, in Akkadian. A typical enty reads:
di-r diri si-y-ku wa-at-ru
This tells us that the sign is read Idi-ri/. Graphically, this sign "looks like" the si
sign ( ) followed by the -sign ( T ). (At least, in this period of cuneifon writing.
Originally, the diri-sign may have had no connecton at all with either the si-sign or the
sign. However, by the Old Babylonian period, when signs were becoming more linear, it
happened to assume a shape looking like the si-sign followed by the -sign.) Because of
this exteral similarity, the Akkadians named this sign "the si of ", that is, si.a. (k). Finally,
the last column gives the Akkadian tanslation, "excess" or "exta".
Copies of these canonical texts have been found all over the Near East, not just in
Mesopotamia. There are also somewhat similar texts, but not of any canonical status, both
frm Mesopotamia and frm outside of Mesopotamia. Their function was the same, to aid
local scribes in their mastery of S umerian (or of some other language).
Some of these non-canonical texts are bilingual, some are tilingual, and some even
quadrilingual. For example, frm Boghazkoy in Asia Minor there are several Sumerian
Akkadian-Hittite tilingual vocabularies. The natve language in Boghazkoy was Hittite;
these texts were designed to help Hittite scribes in learing both Sumerian and Akkadian.
Frm Ugarit, there is a quadrilingual "vocabulary". It has entries in Sumerian, Akkadian,
Souces 275
Hurrian, and then U garitic (written syllabically).
The fnds at Ebla have produced a new, important source of lexical texts of different
kinds. Some are related to the later Mesopotamian tadition; some are independent
creations. Many are monolingual in Sumerian (less fequently in Eblaite). Others are
bilingual, with the Sumerian again on the left and the Eblaite on the right. I some cases,
the pronunciation of the Sumerian is given, using a reduced number of syllabic signs.
Lexical lists ocur among the oldest tablets known. Nissen estmated that some 15%
of the Archaic Texts fom Uruk are lexical texts. One, the "Standard Professions List", i s
well-known fom later copies; the Uruk version lists some 1 00 different professions and
titles.
Lexical lists continued to be used up to the very last stages of cuneiform. Copies of
some of the canonical lexical series are known from as late as Seleucid times, when
Akkadian itself was no longer a spoken language, having been replaced by Aramaic.
The Mesopotamian lexical lists are not always easy to use. Alongside errors of va
rious kinds which have crept in - some due to the normal accidents of textual tansmission,
some due to Akkadian scribes not understanding their originals - there are several con
ventions and abbreviations used by the scribes, which make it diffcult to understand the
texts. Also, these lists should be thought of as a kind of "bare bones" text; there was
undoubtedly a temendous amount of oral information passed on in the Mesopotamia
schools, feshing out these texts. Unfortunately, there are only hints of such oral teaching.
In addition, it is not always easy to recognize the order of entries in the lexical texts.
As Civil has said,
When they attempted to make an inventory of Sumerian words, the native
Mesopotamian scribes faced a problem familiar to any lexicographer in the
frst stages of planning a dictionary: should the entries be organized
thematically, by subjects, or should they be arranged in a serial order based
on graphic or phonological characteristics of the words? One can hardly
speak of planning in the compilation of the Mesopotamian lexical lists as a
whole, since they were the result of a slow process, which lasted for
centuries and answered many different kinds of needs: scribal training,
interpretation of taditional texts, composition of new texts, and, un
doubtedly, a certain amount of simple philological curiosity, spurred on by
the desire of salvaging the words of an extinct language. Nevertheless, the
compilers of each new addition to the taditional lexicographic corpus had to
decide how the enties should be arranged (MSL XIII [ 1971] 3).
Civil ( 1 976) has written a most useful article listing and discussing the most important
Sumerian lexical texts. He also discusses the moder editions of these texts, and the
theoretical principles which need to be observed when attempting to write a Sumerian
dictionary or glossary based on these native sources.
Lexical texts (and the grammatical texts discussed immediately below) were among the
frst cuneiform tablets to be found and published; their importance for the reconstruction of
Sumerian was early recognized. These texts are being systematcally published in a series
entitled Materialien zum Sumerischen Lexikon (MSL). Volume 1 , edited by Benno Lands
berger, appeared in 1937; volume 1 6 appeared in 1976, and several other volumes are still
276 Appendix :
in the process of being prepared.
The role of the lexical lists is usually described as essentially a learing device in the
schools. Mogens Trolle Larsen has discussed the "place of the lexical tadition in the
cognitive scheme ofthe ancient Mesopotamians":
There is no doubt that the lists did function within the scribal world as part of
the basic curriculum in all periods of Mesopotamian history, but it is likewise
possible to understand the lists as sering another purpose, to present a
systematic and ordered picture of the world ( 1987:208-209).
Westenholz says somewhat the same thing:
I many illiterate cultures, an enormously detailed voabulary of plants,
animals, tees, etc. functions as the skeleton of an integrated classifcation
and taxonomy of the known world; and we may see the Sumerian lists of
everything from gods to milk-pots in a similar light as an itemized statement
on the world order, the origin and functioning of which mythology describes
in literary terms (1985:295).
Gramatical texts
The Sumerian language differs from Akkadian not only in vocabulary, but also in
morphology and syntax. These differences led the Akkadian scribes to produce a series of
grammatical texts, in addition to the lexical texts just discussed. The earliest of these are
known as the "Old Babylonian Grammatical Texts", or OBGT. Dating from about 1700-
1 600 BC, these are in the form of paradigms: paradigms of verbs, nominal forms, particles,
etc. The Sumerian is on the left, and the Akkadian (if present) on the right. OBGT VI, for
example, lists over 200 forms of the verb gar, "to place". Lines 124- 1 25 read as follows:
mu-un-gm
mu-gr
This seems to be saying that for the Akkadian scribes, the forms mu-un-gar and mu
gar have the same meaning. This is how moder Sumerologists interpret the two forms, as
a difference in orthography, not in morphology.
Lines 142- 143 of the same tablet read:
ba-an-gar
ba-gar
The Akkadian scribes interpreted the difference in the Sumerian conjugation-prefx as
a difference in the Akkadian stem: mu-un-gar was tanslated by the B-preterite ikun, but
ba-an-gar by the Bt-preterite itakan (there is some reason to assume that itakan is a Bt
preterite, not aB-perfect).
A later series, the "Neo-Babylonian Grammatical Texts" (NBGT), dates frm about
the sixth century BC. These texts are organized according to morpheme: A Sumerian
morpheme is glossed by an Akkadian equivalent. For example, NBGT I, line 153, reads:
Souces 277
The tablet is equating the Sumerian comitatve case-marker da with the Akkadian
preposition qadu, meaning "with".
Lines 405-408 of the same tablet read:
The Sumerian cohortative modal-prefx g, and three morphophonemic alterants of the
desideratve modal-prefxb, are all "tanslated" as the Akkadian desiderative-marker lu.
The NBGT texts occasionally add scribal comments or annotations, in both Sumerian
and Akkadian. Examples include the Sumerian word AN-TA, "prefx", and the Akkadian
expression a iten, "singular". These kinds of annotations do not appear in OBGT.
OBGT and NBGT are rather extensive. One might think that these texts could furish
a key to Sumerian morphology. Unfortunately, it is not so. These texts are all relatively
late. They represent Akkadian-speakers
'
understanding of Sumerian. However, these
Akkadian-speaking scribes did not always understand Sumerian grammatical categories
and distinctions. For example, in the passages fom OBGT cited above, a difference in
Sumerian conjugation-prefx was equated with a distncton in Akkadian stem. It is
diffcult to say how accurate an equation this is. O, Black has pointed out that OBGT V
makes a consistent distinction between the frst person suffx lenl, written -en, and the
second person sufx lenl, written --en ( 1 984:7). Black thinks that this might indicate a
difference in pronunciation, but more likely it is a purely graphic distinction.
Regarding the Sumerian and Akkadian grammatical ters which sometimes occur as
annotations, Black has also said that there is
a growing body of evidence that the scribes responsible for intoducing the
grammatical terms into the grammatical analysis texts sometimes misun
derstood their meaning (or misunderstood the texts into which they were
intoducing them) ( 1984:90).
There is also a certain amount of systematization and schematization in these tablets.
But at the same time, there is no unifed method of organizing the data. As mentioned
above, OBGT has over 200 lines of g, but it is not always easy to follow the principles by
which these forms are organized. Other OBG T texts follow their own organization.
Like the lexical lists, there also ocur mechanical errors due to problems of text
transmission. There are also scribal conventions and abbreviations. And as mentioned
earlier, there was undoubtedly a large oral component which accompanied the study of
these texts, a component which is no longer accessible. These prblems (and others) mean
that although these grammatical texts are a font of useful inforation, this information
cannot be used uncritically. These texts cannot be viewed as an exact refection of
Sumerian of the Ur III or earlier periods. Black has said that "in some cases it seems that
we know Sumerian better than the compilers, or copyists, of our texts" ( 1 984:7). This is
especially true regarding the Sumerian aspectual differences, and also regarding Sumerian
278 Appendix :
causative sentences, a type of sentence heavily dealt with in OBGT. (Similarly, Black
believes that certain Akkadian forms seen in OBGT were "especially concocted to set
against certain Sumerian forms" [ 1984: 29]. )
Thus, although Jacobsen i s undoubtedly correct in saying that the OBGT
constitute without question the most important single group of sources both
for the history of grammatical studies generally and for our understanding of
Sumerian grammar specifcally so far known ( 1 956: 1 *),
these texts must be approached with caution. Jacobsen adds that "the immensity of the
number of problems raised, and the relative insuffciency of our present knowledge of
Sumerian becomes only too clear as one approaches the texts in earnest" ( 1 956: 2*).
As the above quotes exemplify, scholars vary in their estimation of the worth of these
texts in reconstructing Sumerian grammar. Jacobsen, for example, sees them as extremely
important; Kecher, on the other hand, in his study of the conjugation-prefxes containing
an fmf element, found them to be of little value ( 1985: 34).
Finally, Civil has pointed out:
To my knowledge, the fundamental question: are the grammatical texts des
criptve or prescriptive? has never been formulated . . . at least in print, al
though there is widespread skepticism about their descriptive adequacy
( 1986: 72).
For a long time, it was believed that OBGT texts represented mankind's frst attempt to
formulate "paradigms"; these texts antedate the grammatical studies of the Indians by over a
millennium. However, what may well represent the fst attempt known to organize verbal
forms into a paradigm is now found among the Ebla tablets. From a rather variegated
monolingual lexical text (TM. 75. G.2260), the following lines appear:
(line 1 2) : in-na-sum
I-na-sum
nU-l-na-sum
h
i-na-sum
ba-til
nu-til
in-til
h
i-til
For several reasons, one might be hesitant about considering these lines to be a
"paradigm". But they do seem to indicate that the "mind" of the scribe(s) was heading in
that direction. And this tablet antedates the OBGT texts by some 500 years.
The function of the lexical texts (and the smattering of what might be called a
"grammatical text" just discussed) at Ebla was similar to the function of the lexical and
grammatical texts found throughout Mesopotamia and the Ancient Near East in general: to
enable scribes to master Sumerian. For the Eblaite scribes, however, Sumerian was a
language still being spoken. For the Akkadian scribes of the Old Babylonian period,
Sumerian had virtually ceased to exist as a spoken language, and was only a language of
the schools.
Souces 279
OBGT and NBGT were published in MSL IV ( 1956). These texts were prefaced by
Jacobsen with a discussion of the Sumerian verbal system as refected in these texts.
Jacobsen has also written a very interesting aricle intended for a more general linguistic
audience, discussing the system of paradigms seen in OBGT and NBGT ( 1 974). Black
( 1 984) has written a book especially on these grammatical texts, and on the philosophy of
language which they represent; this work also has much incidental discussion of various
aspects of Sumerian morphology.
Very recently, Civil et al have published some "Middle Babylonian Grmmatical
Texts" ( 1 986). These have not yet been fully studied.
Syllabic S umerian
I addition to the "standard" or "normal" orthography and spelling of Sumerian, there
is a certain amount of what is called "syllabic Sumerian" or "phonetic Sumerian". Standard
Sumerian is written using a combination of logographic and syllabic signs. Syllabic
Sumerian, however, is written using only syllabic signs. For example, the standard Su
merian orthography for a locative phrase, kalam-ma, is written in syllabic Sumerian as ka
la-ma.
There are not a great deal of texts in syllabic Sumerian; they are all relatively late.
Interestingly, not many syllabic texts come from the Mesopotamian heartland; they are
mostly fom norther Babylonia, or farther afeld. The practice prbably originated in the
scribal school system as a device for the scribes to cope with the difculties of standard
Sumerian orthography'. In the case of syllabically-written incantations and liturgical texts,
the purpose was probably to aid in correct recitation.
Unlike the lexical and grammatical texts discussed above, there is no standard or
canonical system of writing syllabic Sumerian; it varies to some degree from text to text.
Much of it was prbably prduced on an ad-hoc basis, to deal with particular texts.
Since syllabic Sumerian is an attempt to reproduce spoken Sumerian, it should reveal
some of the features not shown in the normal orthography. Thus, one might think of it as
another key to unlocking Sumerian morphology. Unfortunately, it is very difcult to
understand syllabic Sumerian, even more difcult than it is to understand Sumerian in
standard orthography. The reason is precisely because standard Sumerian masks certain
phonetic prblems, such as morphophonemic alteration, contaction, assimilation, etc.
When such phenomena actually show up in syllabic Sumerian, it is often diffcult to
untangle the forms. Even in cases where the same text is preserved once in standard
orthography and once in syllabic orthography, the phonetic relationship between the two is
not always easy to see.
A relatively simple case is the writing at-ta, for standard an-ta, "fom the sky", or
"frm above". Should it be assumed that the standard Sumerian was also pronounced
latta/, and that the written form an-ta is a morphographemic or historical writing? If so,
should the an-sign be transliterated by an at-value? Perhaps in early Sumerian, the word
was indeed pronounced lanta/, but an assimilation took place in later Sumerian, prducing
latta/. How can this change be dated? On the other hand, perhaps such a writing as at-ta
refects the Akkadian assimilation of nasals, and doesn
'
t say anything about Sumerian.
280 Appendix :
Even in this one simple instance, one can think of several variables which must be
taken into account. But consider an even more complicated case. From the root h
meaning "to be confused", there appears a man form in standard orthography as ba-bir
bir-re (this shows formation of the mari by means of both reduplication and the mari-suf
fx .e). This appears in syllabic orthography as ba-bi-ib- presumably representing
Ibabibre/. Does this mean that the writing in standard orthography, ba-bir-bir-re, should
also be understood as representing Ibabibre/, and that this is a morphographemic or
historical spelling? How should it be transliterated?
Such examples illustrate the extent to which phonetic processes are masked by
standard orthography, and they show the diffculty in interpreting the syllabic forms. And
since syllabic Sumerian varies to some degree from text to text, it is diffcult to generalize
about what is seen. All syllabic texts are rather late, fom the Old Babylonian period or
after. They thus refect a stage when Sumerian was no longer a spoken language, so to
some degree the phonetic differences that appear may be conditioned by the Akkadian
language of the scribes.
Very recently, however, there have been found at Ebla syllabically-written versions of
lexical lists. Civil has called their existence "a most unexpected surprise which opens a
new chapter in the understanding of the earliest lexical compilations and provides
phonological data for the oldest stages of Sumerian" ( 1 982: 1 ) . These syllabically-written
lexical texts are very difcult to interpret, and only preliminary work has been accom
plished.
Much more work, in general, remains to be done on syllabic Sumerian. The two
examples given above show the kinds of information which such texts can provide. A
more thorough investigation might help in solving some of the perplexing problems
encountered in the Sumerian writing system.
The Emesal dialect of Sumerian is written in a mixture of standard (i .e. , mostly
logographic) Sumerian and of syllabic Sumerian. For example, the word for "lady" in
Emesal is gaan, corresponding to Main Dialect nin. Sometimes Emesal texts simpl y use
the same nin-sign; it is assumed that the "reader" will know enough to render the nin-sign
as the Emesal equivalent gaan. More frequently, however (at least, with this particular
word), the word is spelled syllabically, gf-a-an.
An Emesal vocabulary has also been preserved, some 1 77 lines long. It gives the
Emesal form in the far left-hand column; the Main Dialect form; and an Akkadian
translation in the far-ri ght column. A typical example is line 96:
This tells us that the adjective "wide", Akkadian rapu, which is written with the dagal-sign
in Main Dialect, is I damall in Emesal.
Bilinguals
A certain number of "bilingual" texts have come down to us. These are of two main
types. In "interlinear" texts, a line of Sumerian is followed by a line of Akkadian. There
are many incantations of this type. Other texts are written in "parallel columns", with the
Souces 281
Sumerian on the left of a tablet, and the Akkadian on the right.
Bilingual texts are of many different genres. They include incantations, rituals, hymns,
proverbs, letters, and even a few royal inscriptions. Sometimes even more than two
languages are involved. For example, among the texts found at Boghazkoy (a Hittite
speaking area), there is a trilingual poem to the god Ishkur. The text is divided into groups
of four lines. The frst line is Sumerian in standard orthography; the second line is the
same in syllabic Sumerian; the third line is an Akkadian tanslation; the fourth line is a
Hittite translation. This text illustates how syllabic Sumerian was used to help the Hittite
speaking scribes in tying to fgure out the standard Sumerian version.
Most bilinguals are rather late, from the later Old Babylonian period on. Some may be
earlier; for example, there are a couple of Old Babylonian copies of Old Akkadian bilingual
royal inscriptions. One is a bilingual inscription of Sargon, written in parallel columns on
the back of a statue of Sargon; this was re-copied in Old Babylonian times.
Such terms as "bilingual" or "tilingual" are mostly used to refer to single tablets
which have writing in more than one language. There are other cases where a Sumerian
version of a text is present on one tablet, and an Akkadian tanslaton on another tablet.
(Without both versions, it might never be known that the Akkadian was a tanslation.) The
term "bilingual" is occasionally used to refer to such "parallel versions".
One of Shulgi
'
s royal inscriptions has been presered in both a Sumerian version and
an Akkadian version. But neither appears to be a tanslation of the other; rather, they seem
to be independent compositions commemorating the same event, the building of a temple.
Such texts (there are few) are helpful in elucidating the relationship between Sumerian and
Akkadian during the Ur III period.
Other sources
There are a few other sources to aid in reconstcting Sumerian. For instance, there
are Akkadian commentaries to certain genres of text, such as medical texts and omina.
None of these are as extensive or as helpful as the four groups outlined above.
PQQ0u0X 3
Glossa
There is no standard sign-list for Sumerian. However, the sign-lists of Borger and
Labat, even though based on Akkadian values, are still useful for Sumerian. Labat is
especially helpful for the study of the palaeography of the signs, that is, the variation in
their basic shape throughout time. Borger provides some information about Sumerian
grammar, and also a certain amount of bibliographic material. The two volumes of
Ellermeier which have appeared to date are useful in sorting out inconsistencies in
published transliterations.
At the moment, there is no up-to-date dictionary of Sumerian. For a number of years,
the University of Pennsylvania has been preparing just such a project, the Pennsylvania
Sumerian Dictionary (PSD; cited in the Bibliography under Sjoberg [ 1 984]). As of this
writing, the only volume which has appeared so far is for the letter "B". The glossary of
Delitzsch, although dating to 1 914, is still one of the most useful single-volume dic
tionaries, although not the easiest to use. There are a few glossaries to specialized bodies of
texts. One of the more recent is Behrens and Steible ( 1983). Unfortunately, even semi
serious lexicographical work usually means looking at the glossaries and indices of many
different text editions. For the beginning student, perhaps the most useful of these is
Sollberger ( 1 966). Professional Sumerologists keep very large and detailed fles on
Sumerian words. The core of the PSD, for example, is
A
ke Sjoberg's collection of over
500,000 dictionary entries, which he started in 1 949.
a (Text 1 )
arm, strength (Text 22)
a . . . I to dedicate a votive object (Text 4)
ab (Text 17)
abzu apsu (Text 14)
ad (Text 1 9)
ad-da father (Text 1 9)
Adab Adab (GN) (Text 20)
g (Text 8)
igi
2
(Text 8)
alam statue (Text 1 5)
am (Text 1 1 )
ama mother (Text 1 5)
amar young bull (Text 1 3)
Amar-
d
Zuen Amar-Sin (PN) (Text 1 3)
an (Text 1 8)
An An (DN) (Text 6)
an heaven (Text 1 )
283
2
8
4
An-nu-ni-tum
an-ub corner
arad servant
a (Text 22a)
ba (Text 1 0)
Annunitum (DN)
(Text 1 0)
(Text 20)
ba to give as a gift (Text 21 c)
ba
6
(Text 1 1 )
ba-al to dredge (Text 5)
Appendix 3
(Text 1 8)
Ba-ba
6
Baba (DN) (Text 1 1 )
Ba-ba
6
-nin-am Babaninam (PN) (Text 1 1 )
bad city wall, rampart, fortifcation (Text 3)
barag (bara dais (Text 6; Discussion, Lesson 1 5)
barag-sig9-g pedestal (Text 15)
be (Text 9)
bi (Text 9)
bi (Text 1 2)
bi ! (Text 1 2)
Bi l-g-mes Gilgamesh (DN) (Text 1 2)
bU
6
to tear out, to uproot (Text 1 5)
da (Text 1 2)
da side (Text 1 0; Discussion, Lesson 1 2)
Da-da Dada (PN) (Text 21 b)
dab
S
to hold (Text 1 1 )
dadag to be clean, pure (Text 19)
dam wife, consort (Text 1 8)
digi god (Text 6)
di m
(Text 1 2)
di m to fashion, fom1 (Text 1 1 )
du to build (Text 1; Text 1 1 )
dub tablet (Text 19a)
dub-sar scribe (Text 19a; Discussion, Lesson 21 b)
dumu son (Text 9)
dur bond (Text 1 2a)
(Text 1 2)
house (Text 1)
e (Text 22)
E-dur-an-ki Eduranki (TN) (Text 1 2a)
.-kur Ekur (TN) (Text 8)
.-ag4-g-pad-da Eshagepada (TN) (Text 19)
E-temen-ni -guru3 Etemenniguru (TN) (Text 9)
eb (Text 15)
en (Text 2)
en lord (Text 9)
en priest, priestess (Text 1 7)
Glossay
En-di m-gig Endimgig (ON) (Text 1 2)
En-erin
T
nun Enerinnun (ON) (Text 5)
En-ki Enki (DN) (Text 14)
En-lil Enlil (DN) (Text 5)
ensi
2
city governor, local ruler (Text 20)
er (Text 22a)
erin
2
(Text 5)
g (Text 2)
gal great (Text 19; Discussion, Lesson 7)
g (Text 1 6)
g (Text 20)
gi (Text 2)
g to return, to restore (Text 9)
g
!
-na true, correct; standard, certifed (Text 21 a)
gi-na to standardize, to certify (Text 21d)
gig (Text 1 2)
gg (Text 22)
gg-za throne (Text 22)
gg-za-Ia (kind of offcial) (Text 22)
gub to stand; to make stand, to plant (Text 6)
gudug (gudu
4
) (kind of priest) (Text 19; Text 1 1 )
guru3 (Text 9)
M (Text 8)
gal (Text 19)
getin vine, wine (Text 1 6)
getug3 ear, intelligence (Text 22)
gi
6
(Text 17)
g
!6
-Ql giparu (Text 1 7)
gi r (Text 1 1 )
gg (determinative preceding objects of wood) (Text 6)
b (Text 1 2)
I-a-b-me-er Hashamer (PN) (Text 22a)
I-ba-Iu
5
-g Habaluge (PN) (Text 20)
I-la-
d
Ba-ba
6
Halababa (PN) (Text 19a)
b (Text 1 5)
bi (Text 1 1 )
bi-
li
attraction; headdress, wig (Text 1 1 )
! (Text 21 b)
i oil, fat (Text 1 6)
-bi -
d
Zuen Ibbi-Sin (PN) (Text 21b)
i-nun butter (Text 1 6)
ib (Text 1 2)
2
8
5
28
6
id (Text 1 9)
i d (h) river, canal (Text 5)
i (Text 22)
im (Text 1 5)
in (Text 7)
Inanna Inanna (DN) (Text 2)
iq (Text 1 9)
ir (Text 1 4)
is (Text 22a)
Appendix 3
IS-ku-un-
d
Zuen Ishkun-Sin (GN) (Text 22a)
isib (kind of priest) (Text 1 9)
ka
(Text 5)
kalag (kala) to be mighty (Text 2)
kalam land (Text 17)
kar quay, pier; market place (Text 17)
Kar-zid-da Karzida (GN) (Text 17)
ke
4
(Text 1 ; Discussion, Lesson 1 )
ki (determinative following GNs) (Text 1 )
ki place, eanh (Text 6; Text 1 2a; Text 1 4)
ki . . .

2
to love (Text 8)
Ki-en-gi Sumer (GN) (Text 2)
Ki-lul-la Kilula (PN) (Text 22)
Ki-uri Akkad (GN) (Text 2)
kiri
6
garden (Text 6)
kisib cylinder-seal (Text 22)
ku (Text 22a)
kug Gl) bright, pure, holy (Text 1 7; Discussion, Lesson 1 8)
kur mountain; highland; foreign land (Text 5)
kur to change (Text 1 5)
kur
s
C!
s
) (Text 1 2)
kur
9
(ku
4
) to enter (Text 17)
la (Text 4)
la (Text 7)
la to hold, to lift, to carry (Text 22)
Lagas Lagash (GN) (Text 22)
lal honey (Text 1 6)
lam (Text 22)
Lamar DN (Text 1 1 )
le (Text 1 3)
li (Text 1 1 )
l (Text 1 7)
lil air, wind (Text 5)
limmu
2
four (Text 1 0)
li man (Text 7)
Iu
s
(Text 20)
Glossay
lugal king (Text 1 ; Discussion, Lesson 7)
l ul (Text 22)
ma (Text 1 )
ma-da land (Text 1 9)
ma-na mina (Text 21 a)
man to be splendid, magnifcent (Text 6)
mar (Text 1 9)
Mar-tu the Amorites; the West (Text 1 9)
me (Text 1 5)
mes (Text 22)
Mes-lam-ta-- Meslamtaea (DN) (Text 22)
me (Text 1 2)
mu (Text 1 )
mu name (Text 1 5)
mu . . . pad (Il) to propose (Text 1 3)
Mu-r-iq-Ti-id-ni-im Muriq-Tidnim (ON) (Text 1 9)
mu-sar-ra inscnptIon (Text 1 2)
munus woman (Text 1 1 )
na (Text 1 )
nam (Text 4; Discussion, Lesson 9; Sign-list, Lesson 1 2)
nam . . . kur
S
(ku
S
) to curse (Text 1 2)
Nammu Nammu (DN) (Text 1 )
Nanibgal Nanibgal (DN) (Text 21 b)
Nanna Nanna (DN) (Text 1 )
Nane Nanshe (DN) (Text 1 1 )
ne (Text 6)
ne forces, troops (Text 1 9)
H (Text 1 )
ni (Text 9)
Nibru Nippur (ON) (Text 1 3)
nidba food offering (Text 5)
ni g Cr thing (Text 22; Discussion, Lesson 4; Discussion, Lesson 9)
nin lady, mistress; ' lord' (Text 1; Discussion, Lesson 1 1 )
Nin-gal Ningal (DN) (Text 4)
Nin-gir-su Ningirsu (DN) (Text 1 1 )
Nin-li l Ninlil (DN) (Text 1 9)
Nin-tur-dumu-g1
l Q
Ninturdumugu (PN?) (Text 1 6a)
nir (Text 1 9)
nir-gal prince (Text 1 9)
nitan (nita) man, male (Text 2)
nu (Text 1 8)
287
288 Appendix 3
nu not (Text 1 6)
numun seed; offspring, progeny (Text 1 5)
nun (Text 5)
nun prince, noble (Text 1 6)
pad (w) to fnd, call, reveal (Text 1 3)
nl (Text 1 7)
ra (Text 5)
re
(Text 6)
re
6
(Text 1 6a)
ri (Text 1 9)
ru to send (Commentary, Lesson 4)
sag head (Text 9)
sag slave (Text 21 c)
Sag-
d
Nanna-zu Sagnannazu (PN) (Text 21 c)
sag-ls supporter, sustainer, patron (Text 1 3)
sagi (kind of priest) (Text 21 c)
sar (Text 1 2)
sar to write (Text 1 9a)
sig
9
(si) to be narrow (Text 1 5)
sikil to be pure, clean (Text 6)
sipad (sipa) shepherd (Text 1 9)
siskur
2
sacrifce (Text 1 6)
su (Text 1 1 )
sud
4 1 3)
to be long (Text 1 7)
sag
4
(s) heart (Text 1 8)
sag
s
(g
6
) to be good (Text 22)
S
ara
2
Shara (DN) (Text 1 9)
se (Text 4)
silig to cease (Text 1 6)
su (Text 1 8)
su hand (Text 1 9)
su . . . ur to erase (Text 1 2)
S
u-
d
Zuen Shu-Sin (PN) (Text 1 8)
suI (Text 1 0)
S
ul-gi Shulgi (PN) (Text 1 0)
ta (Text 1 7)
temen (Text 9)
ti (Text 1 9)
til to put an end to (Text 1 5)
ti l
Ct)
to live (Text 4; Commentary, Text 22)
tu (Text 1 9)
tur (Text 1 8)
tur (Text 1 6a)
! (Text 1 1 )
! and (Text 1 6)
ub (Text 1 0)
ud ) day (Text 1 2)
ul remote, distant (Text 1 7)
un (Text 17)
Unug (Unu) Uruk (ON) (Text 9)
ur man, warrior (Text 1 )
Ur-ba- ? Urba? (PN) (Text 22)
Glossay
Ur-
d
Lamar Ur-Lamar (PN) (Text 1 9a)
Ur_
d
Nammu Ur-Nammu (PN) (Text 1 )
Ur-
d
Nanibgal Ur-Nanibgal (PN) (Text 21 b)
Ur-
d
Nin-gr-su Ur-Ningirsu (PN) (Text 1 1 )
ur (Text 1 2)
Uri (Text 2)
Urim
s
(Uri
s
) Ur (ON) (Text 1 )
us (Text 1 3)
utu sun (Text 17)
Utu Utu (DN) (Text 4a)
zabar bronze (Text 1 1 ; Discussion, Lesson 20)
zabar-dab
S
(kind of offcial) (Text 1 1 )
zid (zi) effective, true (Text 1 7)
zu (Text 1 3)
zu your (Text 21 b)
zu to know (Text 21 c)
Zuen Zuen (DN) (Text 1 3)
"5" (Text 21 a)
289
Appendix 4
Bibliography
Abbreviations
AfO
AHw
AIUON
AJA
AJSL
ArOr
AS
ASJ
BiOr
CAD
CRAI
GLECS
HUCA
JANES
JAOS
JCS
JNES
MIO
MSL
OA
OIP
Or
PSD
R
RLA
SEb
SEL
TCS
UF
VO
WO
Archiv far Orientforschung
Akkadisches Hand worterbuch
Istituto Universitario Orientale di Napoli -Annali
American Joural of Archaeology
American Journal of Semitic Languages
Archiv Orientalia
Assyriological Studies
Acta S umerologica (Japan)
B i bliotheca Orientalis
Chicago Assyrian Dictionary
Compte rendu de la Rencontre Assyriologique Internationale
Groupe linguistique d'etudes chamito-semitiques -comptes rendus
Hebrew Union College Annual
Joural of the Ancient Near Easter Society
Joural of the American Oriental Society
Joural of CuneifOlID Studies
Joural of Near Eastern Studies
Mitteilungen des Instituts fur Orientforschung
Materialien zum S umerischen Lexikon
Oriens Antiquus
Oriental Institute Publications
Orientalia
Pennsylvania Sumerian Dictionary
Revue d'Assyriologie
Reallexikon der Assyriologie
Studi Eblaiti
Studi Epigrafici e Linguistici
Texts from Cuneiform Sources
U garit-Forschungen
Vicino Oriente
W el t des Orien ts
WZKM Wiener Zeitschrift far die Kunde des Morgenlandes
ZA Zeitschrift fOr Assyriologie
Works Cited
van Aalderen, C.T. 1 982. "Some observations on ergativity and Sumerian". Orientalia
Lovaniensia Perodia 1 3: 25-44.
7V J
292 Appendix 4
Attinger, Pascal. 1 985. "Les prefxes absolutfs de la premiere et de la deuxieme person ne
singulier dans les formes ma ergatives". Z 75: 1 61 - 178.
Bauer, Josef. 1975-76. "Zur /dr/-Phonem des Sumerischen". W08: 1 -9.
Behrens, Hermann and Horst Steible. 1983. Glossar zu den altsumerischen Bau- und
Weihinschriften. Wiesbaden: Franz Steiner Verlag. Freiburger Altorientalische Stu
dien 6.
Biga, Maia Giovanna. 1980. "Bilinguismo sumerico-accadico nell'epistolografa sumerica
e paleoaccadica". V03: 1 21 - 1 29.
Biggs, Robert D. 1974. Inscriptions fom Tell Abu Salabikh. Chicago: University of
Chicago Press. OIP 99.
Black, Jeremy. 1984. Sumerian Grammar in Babylonian Theor. Rome: Pontfcal Bib
lical Institute. Studia Pohl, Series maior 12.
Boisson, Claude. 1988. Constraintes typologiques sur le systeme phonologique du Su
merien (to appear).
Borger, Rykle. 1978. Assyrisch-babylonische Zichenliste. Kevelaer: Verlag Butzon &
Bercker. Alter Orient und Altes Testament 33.
Butz, K. 1973-74. "Konzentationen wirtschaftlicher Macht im Konigreich Larsa: Der
Nanna-Ningal-Tempelkomplex in Ur". WZKM65-66: 1 -58.
Canby, Jeanny Vorys. 1987. "A Monumental Puzzle: Reconstucting the Ur-Nammu
Stela". Expediton 29/ 1 : 54-64.
Christian, Viktor. 1961. Die Herkunft der Sumerer. Vienna.
Civil, Miguel. 1973a. "The Sumerian Writing System". O42: 21 -34.
1973b. "From Enki
'
s Headaches to Phonology". JNS 32: 57-61 .
1976. "Lexicography". AS 20: 123- 157.
1982. "Studies on Early Dynastic Lexicography. I". OA 21: 1 -26.
1985. "On Some Texts Mentioning Ur-Namma". Or 54: 27-45.
Civil, Miguel, Oliver R. Gurey, Douglas A. Kennedy. 1986. Materials for the Sumerian
Lexicon: Suplementry Series 1 . Rome: Pontifcal Biblical Institute.
Clauson, Sir Gerard. 1973. "Philology and Archaeology". Antquity 47: 37-42.
Cooper, Jerrold S. 1973. "Sumerian and Akkadian in Sumer and Akkad". O42: 239-
246.
Dandamayev, Muhammad A. 1982. "The Social Position of Neo-Babylonian Scribes".
in: Klengel, Horst (ed), Gesellschaft und Kultur im alten Vorderasien. Berlin: Aka
demie-Verlag. 35-39.
Delitzsch, Friedrich. 191 4. Sumerisches Glossar. Leipzig. Reprint: Leipzig: Zentral
antiquariat der deutschen demokratischen Republik. 1979.
Diakonoff, I.M. 1965. Semito-Hamitic Langlges: An Essay in Classifcation. Moscow:
Nauka.
----- 1 976. "Ancient Writing and Ancient Written Language: Pitfalls and Peculiarities in
the Study of Sumerian". AS 20: 99- 1 21 .
----- 1983. "Some Refections on Numerals in Sumerian: Towards a History of Mathe
matical Speculation". JAOS 103: 83-93.
Du Bois, John W. 1987. "The discourse basis of ergativity". Language 63: 805-855.
Bibliography
2
9
3
Dunham, Sally. 1986. "Sumerian Words for Foundation". R 80: 31-64.
Edzard, Dietz Otto. 1971f. "lamt!, mara und feie Reduplikation beim sumerischen Ver
bum". Z 61 ( 1 971 ) : 208-232; Z 62 ( 1 972): 1-34; Z 66 ( 1976): 45-61 .
----- 1975. "Zur 'Worbildung im Sumerischen"'. Z 65: 254-257.
----- 1976. "'Du hast mir gegeben
', 'ich habe dir gegeben
'
: Ober das sumerische Verbum
sum". W08: 1 59- 177.
----- 1976-80. "Keilschrift". RLA 5: 544-568.
----- 1980. "Sumerisch 1 bis 10 in Ebla". SEb 1 : 1 21 - 1 27.
Ellereier, Friedrich. 1979f. Sumerisches Glossar. Band 1, Teil 1 : Sumerische Lautwer
te (2 vols). Gottingen: Selbstverlag Dr. Friedrich Ellermeier.
Ellis, Richard. 1968. Foundation Deposits in Ancient Mesopotamia. New Haven: Yale
University Press.
Fales, F.M. 1984. "A Survey of Two-word Eblaic Enties in the Bilingual Vocabularies".
in: Pelio Fronzaroli (ed), Studies on the Language of Ebla. Florence. Quademi di Se
mitistica 13. 173- 1 87.
Falkenstein, Adam. 1959. Das Sumerische. Leiden: E.J. Brill. Handbuch der Orien
talistik.
----- 1 978. Graratik der Sprache Gudeas von Laga
2
. Rome: Pontifcal Biblical Insti
tute. AnalectaOentalia 28, 29, 30.
Finegan, Jack. 1 979. Archaeological History of the Ancient Middle East. Boulder: West
view Press.
Finkel, Irving L. 1 987. "An Issue of Weights from the Reign of Amar-Sin". Z 77:
1 92- 193.
Foster, Benjamin. 1 982. "Ethnicity and Onomastics in Sargonic Mesopotamia". O 5 1 :
297-354.
Foxvog, Daniel. 1975. "The Sumerian Ergative Construction". O 44: 395-425.
Frayne, Douglas. 1983. "Sulgi, the Runner". JAOS 103: 739-748.
Gadd, C.J. 1971 . "Babylonia c. 21 20- 1 800 B. C. " in: LE.S. Edwards et al (eds), The
Cambridge Ancient History
3
. Cambridge: Cambridge University Press. Vol. 1 , Par
2. 595-643.
Ge1b, 1.J. 1948. "A New Clay-Nail of Hammurabi". JS7: 267-271 .
----- 1956. "The Names of Ex-Voto Objects in Ancient Mesopotamia". Names 4: 65-
69.
----- 1962. "Ethnic Reconstruction and Onomastic Evidence". Names 1 0: 45-52.
----- 1 977. "Typology of Mesopotamian Seal Inscriptions". in: Gibson and Biggs ( 1977)
1 09- 1 21 .
----- 1 979a. "Defnition and Discussion of Slavery and Serfdom". UF1 1 : 283-297.
----- 1 979b. "Household and Family in Early Mesopotamia". in: Edward Lipinski (ed),
State and Temple Economy in the Ancient Near East. Leuven. Vol I: 1 -97.
----- 1982. "Terms for Slaves in Ancient Mesopotamia". in: M.A. Dandamayev (ed),
Societies and Langues of the Ancient Near East: Studies in Honor of LM. Diako
, noff. Warinster: Aris and Phillips. 81 -97.
----- 1987. "Compound Divine Names in the Ur lI Period". in: Frncesca Rochberg-
294 Appendix 4
Halton (ed), Langua,e, Literature, and Histor: Philological and Historical Studies
Presented to Erica Reiner. New Haven: American Oriental Society. American Oriental
Series 67. 1 25- 1 38.
Gibson, McGuire and Robert D. Biggs. 1977. Seals and Sealing in the Ancient Near East.
Malibu: Undena. Bibliotheca Mesopotamica 6.
-----. 1987. The Organization of Power: Aspects of Bureacracy in the Ancient Near East.
Chicago: University of Chicago Press. Studies in Ancient Oriental Civilizations 46.
Gordon, Cyrus. 1987. "Eblaitica". in: Cyrus Gordon et al (eds), Eblaitica: Essays on the
Ebla Archives andEblaite Languag. Winona Lake: Eisenbrauns. Vol. 1 : 19-28.
Gragg, Gene. 1968. "The Syntax of the Copula in Sumerian". in: I.W.M. Verhaar (ed),
The Verb "Be" and its Synonyms. Dordrecht. Vol. 3: 86- 1 09.
----- 1972a. "Sumerian and Selected Afr-Asiatic Languages". in: P. Peranteau (ed), The
Chicago Which Hunt. Chicago: Chicago Linguistc Soiety. 1 53- 1 69.
----- 1972b. "Observatons on Grammatical Variation in Sumerian Literary Texts".
JAOS92: 204-21 3.
----- 1973a. Sumerian Dimensional Infxes. Kevelaer: Verlag Butzon & Bercker. Alter
Orient und Altes Testament, Sondereihe 5.
----- 1973b. "A Class of 'When'
Clauses in Sumerian". JNS 32: 124- 134.
----- 1973c. "Linguistics, Method, and Extinct Languages: The Case of Sumerian". O
42: 78-96.
Grayson, A. Kirk. 1980. "Histories and Historians of the Ancient Near East: Assyria and
Babylonia". O 49: 140- 194.
Green, M.W. 1981 . "The Construction and Implementation of the Cuneiform Writng
System". Visible Language 15: 345-372.
Green, M.W. and Hans I. Nissen. 1987. Zeichenliste der archaischen Texte aus Uruk.
Berlin: Gebr. Mann Verlag. Archaische Texte aus Uruk Band 2.
Haayer, G. 1 986. "Languages in Contact: The Case of Akkadian and Sumerian". in:
H.L.I. Vanstiphout et al (eds), Scripta Signa Vocis. Groningen: Egbert Forsten. 77-
84.
Hallo, William W. 1957. Early Mesopotamian Royal Titles: A Philologic and Historical
Analysis. New Haven: American Oriental Soiety.
1962. "The Royal Inscriptons ofUr: A Typology". HUCA 33: 1 -43.
----- 1963. "Lexical Notes on the Neo-Sumerian Metal Industy". BiO20: 1 36- 1 41 .
----- 1966. "The Cornation of Ur-Nammu". JCS 20: 133- 141.
Hallo, William W. and 1.1.A. van Dijk. 1968. The Exaltation of Inanna. New Haven:
Yale University Press. Yale Near Easter Researches 3.
Heimpel, Wolfgang. 1987. "Gudea
'
s Fated Brick". JNS 46: 205-21 1 .
Horsnell, Malcolm. 1977. "The Grammar and Syntax of the Year-Names of the First Dy
nasty of Babylon". JNS 36: 277-285.
lacobsen, Thorkild. 1953. "The Reign of Ibbi-Suen". Reprinted in: Moran ( 1970) 173-
1 86; 408-41 6.
----- 1956. "Intoduction to the Chicago Grammatical Texts". in: MSL I, 1 *-50*.
Bibliography 295
----- 1957. "Early Political Development in Mesopotamia". Reprinted in: Moran ( 1970)
1 32- 1 56; 366-396.
1965. "About the Sumerian Verb". Reprinted in: Moran ( 1970) 245-270; 430-466.
1968. "Some Sumerian City-Names". JCS 21 : 1 00- 103.
1973. "Notes on the Sumerian Genitive". JS 32: 1 61 - 1 66.
1974. "Very Ancient Texts: Babylonian Grammatical Texts". in: D. Hymes (ed),
Studies in the Histor of Linguistics: Traditions and Paadigms. Bloomington:
Indiana University Press. 41 -62.
----- 1987. The Has that Once . . . Sumerian Poet in Translation. New Haven: Yale
University Press.
----- 1988. "Sumerian Gramma Today". JADS 108: 123- 1 33.
Kaneva, I.T. 1970. "Partciples in Sumerian". MID 16: 541 -565.
Kang, S.T. 1968. A Study of the Sumerian Verb in Bilingual Grammatcal Texts. Cin-
cinnati: Hebrew Union College [dissertation] .
Kienast, Burkhat. 1975. "Zur Wortbildung des Sumerischen". Z 65: 1 -27.
----- 198 1a. "1st das Neusumerische eine lebende Sprache?" CRAI28: 105- 1 1 1 .
----- 1981 b. "Probleme der sumerischen Grammatik". Z 70: 1 -35.
Kramer, Samuel Noah. 1936. "Studies in Sumerian Phonetcs". Ar8: 1 8-33.
----- 1940. Lamentation over the Destruction of Ur. Chicago: University of Chicago
Press. AS 1 2.
----- 1967. "The Death of Ur-Nammu and His Descent to the Netherworld". JCS 21 :
104- 1 22.
----- 1983. "The Ur-Nammu Law Code: Who Was Its Author?" O52: 453-456.
Kraus, ER. 1958. "DLtil.1a. Sumerische Prozessprotokolle und Verwandtes aus der Zeit
der Ill. Dynastie von Ur". BiO 15: 70-84.
Krecher, Joachim. 1967. "Zum Emesal-Dialekt des Sumerischen". in: Heidelberger Stu
dien zum Alten Orient. Wiesbaden: Otto Harrassowitz. 87- 1 1 O.
----- 1967-68. "Die pluralischen Verba fur 'gehen
'
und 'stehen
'
im Sumerischen". WD4:
1 - 1 1 .
----- 1978. "Die Form und der Gebrauch der nominalen Verbalformen und die Determina
tion im Sumerischen". Dr47: 376-403.
1985. "Die Im/-Prifxe des sumerischen Verbums". O 54: 1 33- 1 81 .
----- 1987a. "DU KU
x
(-r) 'einteten
'
, 'eineinbringen"
'
. ZA 77: 7-21 .
----- 1987b. "Morphemeless Syntax in Sumerian as Seen on the Background of Word-
composition in Chukchee". ASJ9: 67-88.
Krispijn, Th. 1982. "Das Verb in den zweisprachigen Inschrifen der Hammurabi
Dynasty". in: G. van Driehl (ed), Zikir umim: Assyriological Studies Presented to
ER. Kraus. 145-162.
Kupper, J.-R. and Edmond Sollberger. 1971 . Inscriptions royales sumeriennes et akka
diennes. Paris.
Kutscher, Eduad Yechezkiel. 1982. A Histor of the Hebrew Langl . Leiden: E.J.
Brill.
Kutscher, Raphael. 1975. Oh Angry Sea (a-ab-ba hu-Iuh-ha): The Histor of a Sumerian
296 Appendix 4
Cong@gational Lament. New Haven: Yale University Press. Yale Near Eastern Re
searches 6.
Labat, Rene. 1 976. Manuel d'epigraphie akkadienne
5
. Paris: Paul Geuthner.
Lambert, Maurice. 1 957. "Les noms du pere en Sumerien". Proceedings of the Twenty
Second Congress of Orientalists. Leiden: E. J. Brill. Vol. 2: 27-29.
----- 1 972-73. "Probleme essentiel dans l'etude du Sumerien: Les prefxes verbaux".
GLECS 1 7: 1 3-21 ; 97- 1 04.
Lambert, W. G. 1 975. (review of Henri Limet, Les legendes des sceaux cassites [ 1 97 1 ] ) .
BiOr32: 21 9-223.
Larsen, Mogens Trolle. 1987. "The Mesopotamian Lukewarm Mind: Refections on
Science, Divination, and Literacy". in: Francesca Rochberg-Halton (ed), Lang.!M
Literature, and History: Philological and Historical Studies Presented to Erica Reiner.
New Haven: American Oriental Society. American Oriental Series 67. 203-205.
Lieberman, Stephen J. 1 977. The Sumerian Loanwords in Old-Babylonian Akkadian.
Missoula: Scholars Press. Harvard Semitic Studies 22.
----- 1 978. (letter to editor). Scientifc Amercan 239/ 1 0: 1 5.
----- 1 979. "The phoneme /0/ in Sumerian". in: Studies in Honor of Tom B. Jones.
Kevelaer: Verlag Butzon & Bercker. Alter Orient und Altes Testament 203. 21 -28.
----- 1 980. "Of Clay Pebbles, Hollow Clay Balls, and Writing: A Sumerian View".
AlA 84: 339-358.
Limet, Henri. 1 975a. "Le morpheme suffxe /-a/ en Sumerien". RA 69: 5- 1 8.
----- 1 975b. "Permanence et changement dans l a toponymie de l a Mesopotamie antique".
in: La Toponymie antique. Actes du Colloque de Strasbourg 1 2- 1 4 juin 1 975. Leiden:
EJ. Brill. 83- 1 1 5.
----- 1 978. "Etude semantique de ma.da, kur, kalam". R 72: 1 - 1 2.
Martiny, Giinther. 1 940. "The Orientation of the Gimilsin Temple and the Palace Chapel".
in: Frankfort, Henri, Seton Lloyd, and Thorkild J acobsen, The Gimilsin Temple and
the Palace of the Rulers at Tell Asmar. Chicago: University of Chicago Press. OIP
43. 92-96.
Michalowski, Piotr. 1 980a. "Sumerian as an Ergative Language". lCS 32: 86- 1 03.
----- 1 980b. "Konigsbriefe". RLA 6: 5 1 -59.
----- 1 987. "Charisma and Control: On Continuity and Change in Early Mesopotamian
Bureacratic Systems". in: Gibson and Biggs ( 1 987) 50-68.
Moorey, P. R. S. 1984. "Where did they bury the kings of the IIIrd dynasty of Ur?". Iraq
46: 1 - 1 8.
Moran, William (ed). 1 970. Toward the Image of Tammuz. Cambridge: Harvard Univer
sity Press.
Nissen, Hans J. 1 985. "The Emergence of Writing in the Ancient Near East". Inter
disciplinary Science Reviews 1 0/4: 349-361 .
----- 1 986. "The archaic texts from Uruk". World Archaeology 1 7/3: 3 1 7-334.
----- 1988. The Early History of the Ancient Nea East: 9000-2000 B. C. Translated by
Elizabeth Lutzeier, with Kenneth J. Northcott. Chicago: University of Chicago Press.
Oberhuber, Karl. 1 982. "Zum 'Passivum' im Sumerischen". in: W. Meid (ed), Sprach-
Bibliography
wissenschaft in Innsbruck. Innsbruck. 1 29- 1 34.
297
Oppenheim, A. Leo. 1971 . (letter to editor). Cr entAntrpology 12: 219-220.
Pettinato, Giovanni. 198 1 . "La pronuncia sumerica dei numeri da 1 a 10 in un testo lessi
cale di Ebla". AIUON41 : 141-143.
Poebel, Ao. 1923. Grunzuge der sumerischen Grammatik. Rostok: Selbstverlag des
Verfassers.
----- 1932. Das apsitionell bestimmte Pronomen der 1 . pers. sing. in den westsemiti
schen Inschriften und im Alten Testament. Chicago: University of Chicago Press. AS
3.
----- 1935. "The Sumerian Genitive Element". AlSL 51 : 145-176.
Postgate, J.A. 1974. "Two Points of Grammar in Gudea". lCS 26: 1 6-54.
Powell, Marvin A. 1978. "Ukubi to Mother ... The Situation is Desperate". Z 68:
1 63-195.
----- 198 1 . "Three Problems in the History of Cuneiform Writing: Origins, Direction of
Script, Literacy". VisibleLanguage 1 5: 41 9-440.
Renger, Johannes. 1967f. "Untersuchungen zum Priestertum in der altbabylonischen
Zeit". Z 58 ( 1 967) 1 1 0- 1 88; Z 59 ( 1969) 104-230.
Roberts, Jack. 1972. The Earliest Semitic Pantheon. Baltimore: Johns Hopkins Univer
sity Press.
Romer, W.H.Ph. 1965. Sumerische "Konigshymnen" der Isin-Zeit. Leiden: E.J. Brill.
----- 1973. "Sumerian". in: J. Hospers, A Basic Bibliogr y of the Semitic Languages.
Leiden: E.J. Brill. 38-45.
----- 1983. Einfuhrung in die Sumerolog
4
. Nijmegen: Katholieke Universiteit Nijme
gen.
Rosengarten, Yvonne. 1 967. Repertoire commente des Signes Presargoniques Sumeriens
de Lagas. Paris.
Salonen, Armas. 1961 . Die Turen des alten Mesopotamien. Helsinki: Annales academiae
scientiarum fennicae.
Schmandt-Besserat, Denise. 1982. "How Writing Came About". Zitschrft f Papy
rlogie und Epigaphie 47: 1 -5.
----- 1986. "The Origins of Writing: An Archaeologist's Perspective". Wrtten Commu
nicaton 3: 31 -45.
Seux, M.-J. 1 967. Epithetes royales akkadiennes et sumeriennes. Paris: Letouzey et Ane.
Sjoberg, Ake and E. Bergmann. 1 969. The Collection of the Sumerian Temple Hymns.
Loust Valley: J.J. Augustin. TCS 3.
Sjoberg, Ake et al. 1984f. The Sumerian Diction1. Philadelphia: The Babylonian
Section of the University Museum.
Snell, Daniel. 1 982. Ledgers and Prices: Early Mesopotamian Merchant Accounts. New
Haven: Yale University Press. Yale Near Easter Researches 8.
Sollberger, Edmond. 1952. Le systeme verbal dans les inscriptions "royales" pre-sargo
niques de Lagas. Geneva: Librairie E. Droz.
----- 1966. Business and Administrative Correspondence under the Kings of Ur. Loust
Valley: J.J. Augustin. TCS 1 .
298 Appendix 4
----- 1972. "Ur III Society: Some Unanswered Questions". CRAI 1 8: 1 85- 1 89.
----- 1973. "Poblems of Translation". O 42: 158- 1 61 .
Steiner, Gerd. 1978. "Der Gegensatz 'eigenes Land
'
: 'Ausland, Fremdland, Feindland
'
in
den Vorstellungen des alten Orients". CRAI25: 633-664.
----- 1980. "The Vocalization of the Sumerian Verbal Morpheme /=ed/ and its Signi
fcance". JS 40: 21-41.
----- 198 1 . "lamty und maru als verbale Kategorien im Sumerischen und im Akka
disc hen". RA 75: 1-14.
Steinkeller, Piot. 1977. "Seal Pactice in the Ur III Period". in: Gibson and Biggs
( 1977) 41-53.
1979. "Notes on Sumerian Plural Verbs". Or48: 54-67.
1984. "The Sumerian Verb lug
x
(LUL)". SELl : 5-17.
1987a. (review of Pennsylvania Sumerian DictionI). lNS 46: 55-59.
1987b. "The Administrative and Economic Organization of the Ur III State: The
Core and the Periphery". in: Gibson and Biggs ( 1987) 19-41 .
----- 1988. "On the Identity of the Toponym LV.SU(.A)". lAOS 108: 197-202.
Stola, R. 198 1 -82. "Zu den sumerischen Entsprechungen des akkadischen Imperativs in
spaten zweisprachigen Texten". Af028: 79-91 .
Strommenger, Eva. 1980. "The Chronological Division of the Archaic Levels of Uruk
Eanna VI to IIIIIl: Past and Present". AlA 84: 479-487.
Sullivan, Brian. 1979. Sumerian and Akkadian Sentence Stucture in Old Babylonian
LiterI Bilingual Texts. Cincinnati: Hebrew Union College [dissertation] .
Thomsen, Marie-Louise. 1984. The Sumerian Language: An Introduction to its Histor
and Grammatical Stucture. Copenhagen: Akademisk Forlag. Mesopotamia: Copen
hagen Studies in Assyriology 10.
Tigay, Jeffey H. 1982. The Evolution of the Gilgamesh Epic. Philadelphia: The Univer
sity of Pennsylvania Press.
Vaiman, A.A. 1974. "Uber die Potosumerische Schrift". Acta Antqua 22: 1 5-27.
van Driehl, G. 1982. "Tablets from Jebel Aruda". in: G. van Driehl (ed), Zikir umim:
Assyriological Studies Presented to F.R. Kraus. 12-25.
Vanstiphout, Herman L.J. 1985. "On the Verbal Prefx lif in Standard Sumerian". RA
79: 1 - 1 5.
Veenhof, K.R. 1986. "Cuneiform Archives. An Intoduction". in: K.R. Veenhof (ed),
Cuneiform Archives and Libraries. Leiden: Nederlands Instituut voor het Nabije Oos
ten.
Walker, c.B.F. 1987. Cuneiform. Berkeley/London: University of Califoria Press and
the British Museum.
Weadock, Penelope. 1975. "The giparu at Ur". Irq 37: 1 01 - 1 28.
Westenholz, Aage. 1985. "An essay on the Sumerian 'Lexical
'
Texts of the Third Millen
nium". O 54: 294-298.
Wiseman, D.J. 1960. "The Goddess Lama at Ur". Iraq 22: 166- 1 71 .
Woolley, Sir Leonard. 1982. Ur 'of the Chaldees
'
. Revised and updated by P.R.S.
Moorey. Ithaca: Cor ell University Press.
Biblography 299
Yoshikawa, Mamoru. 1968a. "The Ma and Iamu Aspects in the Sumerian Verbal
System". O37: 401 -41 6.
----- 1968b. "On the Grammatcal Function of -E- of the Sumerian Verbal Suffx -E-DEI-
E-DA(M)". JNS 27: 251 -261 .
1 974. "The Maru-Conjugation in the Sumerian Verbal System". O 43: 17-39.
1 979. "Verbal Reduplication in Sumerian". ASJ 1 : 99- 1 1 9.
1980. "On the Alleged 'Ueberhlngende Vokale
'''
. ASJ2: 1 88- 1 95.
198 1 a. "The Sumerian Verbal Prefx al-". CRAI28: 66-71 .
1 981b. "Plural Expressions in Sumerian Verbs". ASJ3: 1 1 1- 1 24.
Yushu, Gong. 1 987. "ErgativiUit und das Sumerisch". Joual of Ancient Civilizatons
2: 85- 120.
Zakar, Andras. 1 971 . "Sumerian-Ural-Altaic Affnities". Cur ent Antrpology 12:
21 5-225.
Zettler, Richard. 1 977. "The Sargonic Royal Seal: a Consideration of Sealing i Meso
potamia". in: Gibson and Biggs (1977) 33-39.
Zhi, Yang. 1987. "The Name of the City Adab". Joual of Ancient Civilizatons 2:
1 21 - 1 25.
Concordance of texts used
This concordance lists all the texts used in the Lessons. For each text, the following
information is given: number according to Hallo's catalogue (if available); publicaton of
cuneiform; publication of tansliteration and tanslation in SAKI; publication of tanslation
in IRSA; location of photograph. The following abbreviations are used:
Cuneiform
BE Babylonian Expedition of the University of Pennsylvania, Series A, Cuneiform
Texts. Philadelphia.
BE 1 1 896 (Hilprecht)
BIN Babylonian Inscriptions in the Collection of James B. Nies. New Haven.
BIN 2 1 920
CT Cuneiform Texts fom the British Museum. London.
C 5 1 898
C 21 1905
CT 36 1 921
MDP Memoires de la Delegation en Perse. Pais.
MP 6 1 905
OIP Oriental Institute Publications. Chicago.
OIP 14 1 930
PBS Publications of the Babylonian Section. The Museum, University of
Pennsylvania. Philadelphia.
PBS 1 3 1 922
PBS 1 5 1 926
R Rawlinson, H. The Cuneiform Inscriptions of Wester Asia. London.
300 Appendix 4
I R 1 861
UET Ur Excavations. Texts. London.
UET 1 1 928
UET 3 1 937
UET 8 1 965
UVB Vorlaufger Bericht iiber die von der Deutschen Forschungsgemeinschaft in Uruk
Warka unterommenen Ausgrabungen. Berlin.
UVB 10 1 939
UVB 1 2/ 1 3 1 956
VAS Vorderasiatische Denkmaler. Berlin.
VAS 1 1 907
YOS Yale Oriental Series. New Haven.
YOS 1 1 91 5
Secondary Literature
Hall H.R. Hall, A Season's Work at Ur. London, 1 930.
Hallo W.W. Hallo, "The Royal Inscriptions of Ur: A Typology". HUCA 33 ( 1 962) 1 -
43.
HSA L. King, History of Sumer and Akkad. London, 1 91 0.
IRSA E. Sollberger and J. -R. Kupper, Inscriptions royales sumeriennes et akkadiennes.
Paris, 1 97 1 .
SAKI F. Thureau-Dangin, Die sumerischen und akkadischen Konigsinschriften.
Leipzig, 1 907.
Several individuals and publishers have either supplied photographs, or provided
permission to reproduce copyrighted material. I would expressly like to thank the fol
lowing: The British Museum, for supplying photographs of the drawing of the ziggurat in
Lesson 1 ; of the drawing of the ziggurat in Lesson 9; of the pedestal in Lesson 1 5; of the
Old Akkadian seal in Lesson 21 ; of texts CT 21 , 2: 90004; CT 21 , 3: 9001 5; CT 21 , 6:
891 26; UET 1, 55. The Pierpont Morgan Library, for the photographs in Lesson 7 of the
cuneiform tablet and the canephore fgures. The University Museum of the University of
Pennsylvania for the photographs of the stela in Lesson 1 and the stamp seals in Lesson 2.
Jack Finegan, for the photograph of the ziggurat in Lesson 1 (reproduced courtesy of
Westview Press). Corn ell University Press, to reprint from Ur 'of the Chaldees': A
Revised and Updated Edition of Sir Leonard Woolley's Excavations at Ur by P. R. S.
Moorey 1 982 The Estate of Sir Leonard Woolley and P. R. S. Moorey. The Pontifcal
Biblical Institute, for several quotations from Sumerian Grammar in Babylonian Theory by
Jeremy Black 1 984. Any inadvertent omission of acknowledgments will be corrected in
the future.
Textual Concordance
Text 1 Ur-Nammu 8
UET 1 , 40
Drawing of ziggurat: Courtesy of British Museum. Reproduced in Woolley
Text 2
Bibliography
( 1 982) 148
Photograph of ziggurat: Courtesy of Jack Finegan. Reproduced in Finegan
( 1 979) 5 1 . Reproduced courtesy of Westview Press
Photograph of stela: Courtesy of University Museum, University of
Pennsylvania
Ur-Nammu 7i
CT 2 1 , 3: 9001 5
SAKI 1 86d
301
Photograph: Courtesy of British Museum. Reproduced in RSA, facing page
280
Text 3
3a
3b
Photographs of stamp seals: Courtesy of University Museum, University of
Pennsylvania
Ur-Nammu 9i
CT 2 1 , 2: 90004
Photograph: Courtesy of British Museum. Reproduced in Rall 1 06
CT 21 , 2: 90009
SAKI 1 86b
IRSA IIIAlc
Text 3c Ur-Nammu 7i
CT 2 1 , 3: 90006
SAKI 1 86d
Text 4 Ur-Nammu 3 1
UET 1 , 34
Text 4a Ur-Nammu 1 1
CT 21 , 5: 90001
SAKI 1 96e
Text S Ur-Nammu 23i
UET 1 , 46
Text 6 Ur-Nammu 5ii
UET 1 , 41 a
IRSA IIIAlj
Text 7a Ur-Nammu 3iii
Hallo, in: Ancient Mesopotamian Art and Selected Texts. New York: The
Pierpont Morgan Library ( 1 976), p. 22
Photograph: Courtesy of The Pierpont Morgan Library. Reproduced in ibid, p.
1 7
Photograph of canephore fgurines: Courtesy of The Pierpont Morgan Library.
Reproduced in ibid, p. 22
Text 7b Ur-Nammu 3ii
CT 21 , 4: 90802
SAKI 1 86g
Text 7c Ur-Nammu Si
CT 21 , 5: 90296
302
SAKI 1 86f
IRSA IIIAlj
Text 8 Ur-Nammu 1 6
BE 1 , 1 21
SAKI 1 88k
Text 8a Ur-Nammu 3i
BE 1, 1 22
SAKI 1 87, note h
Text 9 Ur-Nammu l Oi
C 21 , 7: 90000
SAKI 1 86c
IRSA IIIAld
Appendix 4
Drawing of ziggurat: Courtesy of British Museum; reproduced in Woolley
( 1 982) 234
Text 10 Shulgi 52
VET 1 , 55
Photograph of 30-mina weight: Courtesy of Britsh Museum
Text lOa Ur-Nammu lOi
I R I i 5
SAKI 1 86c
IRSA IIIAld
Text 1 1 Shulgi 29
C 5, 2: 1 221 8
SAKI 194x
IRSA IIIA2u
Photograph: Hallo, in Denise Schmandt-Besserat (ed), The Legacy of Sumer.
Malibu: Undena Publications ( 1 976), p. 1 33
Text l l a Ur-Nammu I v
Boson, Aegyptus 1 5 ( 1 935) 420
IRSA IIIAl b
Text 1 2 (Ur-Nammu)
UET 8, 4: 21
IRSA IIIAlk
Text 1 2a Shulgi 20i
PBS 1 5, 42
IRSA IIIA2g
Photograph is of Shulgi 22ii: McCown, Archaeology 5 ( 1 952) 74
Text 1 3a Amar-Sin 2ii
C 21 , ' 24: 90034
SAKI 1 96b
IRSA IIIA3b
Text 1 3b Amar-Sin 2iii
VAS 1 , 26
IRSA llIA3b
Text l 3e Amar-Sin 1
BE 1 , 22
SAKI 196a
IRSA IIIA3a
Text 14 Amar-Sin 5iv
Biblography
Sumer 3 ( 1947) l e, facing p. 236
IRSA llIA3h
Text 14a Vr-Nammu 22i
VET 1 , 45
Text 1 5 Amar-Sin 3i
e 21 , 25f: 908 1 1
SAKI 198d
IRSA IIIA3e
Photograph of pedestal: Courtesy of British Museum. VET 1 , 172, pI. V
Text 1 6 Amar-Sin 1 0
BE 1 , 21
SAKI 1989
IRSA IIIA3f
Text 1 6a Shulgi 46
RO 2 ( 1925) 1 89
IRSA IIIA2f
Text 17 Amar-Sin 1 1
VVB 1 0, pI. 28
IRSA IIIA3d
Photograph: VVB 10, pI. 23a
Text 1 8: Shu-Sin 6i
CT 21 , 28: 90844
SAKI 200b
IRSA IIIA4e
Text 1 8a Shulgi 43
MP 6, 22
SAKI 194y
Text 19 Shu-Sin 9iii
YOS 1 , 20
IRSA IIIA4d
Text 19a Amar-Sin 17
BIN 2, 17
Text 20 Shu-Sin 3
OIP 14, 43
IRSA IIIC1 b
Text 21 a Shu-Sin 17
de Sarzee, Deeouvertes en Chaldee ( 191 2) voI. 2 pI . 26-bis: 5
303
30
SAKI 202e
IRSA IIIA4i
Text 21 b Ibbi-Sin 7i
UT 3, 52
Appendix 4
Photograph of Old Akadian seal: Courtesy of British Museum.
Text 21 c Ibbi-Sin 8iii
PBS 1 3, 5: CBS 1 2570
Text 21 d: Shulgi 5l iii
UT 1 , 287
Photograph: The Illustated Bible DictionI (Tyndale, 1 980), vo!. 3, 1 635
Text 22 Shulgi 47
C 21 , 9: 891 31
SAKI 1 94z
IRSA IIIA2w
Collation: Gadd, Iraq 1 0 ( 1 948) 98 n. 1
Photograph: Courtesy of British Museum. Reproduced in RSA, facing p. 246
Text 22a Ur-Nammu 36
C 21 , 6: 891 26
SAKI 1 88n
IRSA IIIA1 m
Photograph: Courtesy of British Museum. Reproduced i n RSA, opposite p.
246
PQQ0B0X 5
Further Work
The Ur III royal inscriptions studied in this book present only a limited picture of
Sumerian. For example, not all the modal-prefxes occur, and there is only limited use of
the frst and second persons. In order to deepen one's knowledge of Sumerian, it is neces
sary to do two things: read well edited Sumerian texts, and read the most important and/or
recent secondary literature about Sumerian.
More "variety" in grammar occurs in Sumerian literary texts, and for that reason they
are perhaps the most useful texts to read at this stage. However, this is not as easy as it
sounds. Because most literary texts were copied down in the Old Babylonian period or
later, they are often infuenced by Akkadian grammar, and may contain forms which are
simply "wrong" by the normative rules of Classical Sumerian grammar; such wrong forms
can be disconcerting to a relati ve beginner.
To obviate this problem, it is necessary to work through literary texts which are well
edited. However, some Sumerologists are less interested in grammatical matters than other
Sumerologists, and so may not discuss such matters as, for example, a seemingly incorrect
use of a personal-affx. Some scholars are more interested in lexicographical matters, and
may devote seemingly inordinate amount of space for citing all references for particular
words, instead of focusing on the grammar (given the lack of a complete up-to-date
dictionary of Sumerian, thi s is often necessary).
In general, it is always valuable to frst skim through an edition of a literary text, in
order to determine what kind of emphasis the moder-day editor is placing in his
commentary -is it primarily a grammatical commentary, or lexicographic, or stylistic, etc.
When reading through a new text, one will encounter variations of constructions seen
previously, or completely new constructions. These may or may not be discussed in a
commentary. Upon encountering a new form or construction, the frst step is to isolate the
problem, that is, determine where in the grammar the problem lies: is it a previously unseen
modal-prefx, or a strange use of a dimensional-prefx, or a problem in a temporal clause,
etc. At this stage, one must turn to the standard grammars, and to other secondary
li terature.
The most general grammars which can be examined are, in chronological order (these
are further discussed below): Poebel ( 1 923); Falkenstein ( 1 959); Falkenstein ( 1 978
2
);
Romer ( 1983
4
); Thomsen ( 1 984). Thomsen will be of most value to the student; she also
includes references to other secondary literature.
In addition to these more general grammars, there are certain articles, on specifc
aspects of the grammar, which are always worth consulting. These are referred to in
Tomsen, but it is useful to keep photocopies of them at hand. Full references are given in
theBibliography.
moods: J acobsen ( 1 965); Edzard ( 1971 f; Michalowski ( 1 980a); Kienast ( 1 981 b)
conj ugation-prefxes:
J acobsen ( 1965); Gragg ( 1968; 1 972a)
305
306 Appendix 5
dimensional-prefes: Gragg ( 1 973a); Yoshikawa ( 198 1 a)
personal-affes: Michalowski ( 1980a)
iaty - man: Yoshikawa ( 1968a; 1968b; 1974); Edzard ( 1 971 f; Michalowski
( 1980a); Steiner (1981)
plural verbs: Krecher (1967-8); Steinkeller ( 1 979); Yoshikawa ( 1 981 b)
imperatives: Michalowski ( 1980a)
copula: Gragg (1968)
subordinate clauses: Gragg ( 1 973b)
relative clauses: Gragg (1972)
ergativity: Foxvog ( 1 975); Michalowski ( 1980a); van Aalderen ( 1982)
As stated above, in additon to reading more texts, it is necessary to read more about
the language itself. At this point, the student can proft by a systematic reading of
Thomsen. Thomsen cannot practically be used without some previous knowledge of
Sumerian, but this Manual should provide enough background to make Thomsen useable.
The aticles listed above, besides being consulted for reference when an unfamiliar form is
encountered, should also be read in their entirety.
For further bibliography on grammatical subjects, one can consult: Sollberger ( 1952),
with a rather complete bibliography through 1952; R6mer ( 1 973); Falkenstein ( 1 978
2
);
R6mer ( 1 983
4
); and Thomsen ( 1984).
There have been fve modem larger-scale grammars of Sumerian. The frst of these
was Poebel ( 1 923). Poebel was perhaps the frst person to really understand Sumerian
grammar, and to put its study on a sound footing. Although written in a very old-fashioned
style, and although wrong in a number of details, it is still a valuable tool.
Falkenstein published a complete "sketch" of Sumerian in 1 959. This work is often
cited by Sumerologists, but it suffers from being too terse (the section on morphology and
syntax occupies less than thirty pages), and from a rather whole-sale use of concepts and
terms which are more ftting for the description of Indo-Eurpean or Semitic languages. It
also has a rather confusing mixture of synchronic and diachronic description. However,
since Falkenstein
'
s views differ somewhat from those presented in this book, and since his
views are accepted (sometimes only tacitly) by many Sumerlogists, this book should be
worked through.
Falenstein has also produced a detailed grammar of the language of the inscriptions
of Gudea (second edition, prepared posthumously, 1978). Although limited to one group
of texts, this is a very useful work. There are three volumes: ( 1 ) script and morphology;
(2) syntax; and (3) an intoduction to the texts, discussing historical background, iden
tifying divine names, etc. This introduction is handy for identifying proper names oc
curring in other texts. The morphology volume contains elaborate paradigms; these are
quite useful, and are in general more accurate than those in Poebel.
Futher Work 307
Romer ( 1983
4
) is essentially an updated Falkenstein; the section entitled "Einiges zur
Sprache" is less than ffty pages long. However, it has an enormous amount of
bibliographic references -there are 641 footnotes; these references are what make the work
valuable. A new edition is being prepared.
Finally, the most recent grammar is that of Thomsen ( 1984); there is an imporant
review by Jacobsen ( 1 988). Thomsen bases her work on the Old Babylonian literary texts,
because these exhibit the greatest variety of grammatical variation. The danger with this
approach is the fact that in these texts one must always b on guard against Akkadian
infuence. I spite of this methodological criticism, Thomsen contains a balanced,
thorough, and relatively uncontoversial presentation of the grammar, more-or-Iess in the
Falkensteinian tadition. Her book would be difcult for anyone who does not know the
principles of cuneiform script, but for those who already know some Sumerian, it will
serve as a standard reference for some time.
This Manual will be followed by a second volume consisting of extacts frm Inanna
'
s
Descent, with a commentary focussing on the grammar of the text.
PQQ0u0X 6
Topical Index
The index is designed to tie together information scattered throughout this book, in
order to permit quick reference and review. Only signifcant discussions are listed. Re
ferences are to pages.
ablative case 177, 197
absolute case 34
abstract nouns 67, 1 05, 251
adjectives 50, 82, 84, 97, 208, 209, 217
adnominal cases 54
adverbial cases 54, 68, 138
Amar-Sin 1 54, 209
amissability 3 1 , 83, 1 21
animacy 32, 39, 1 34, 173
anticipatory genitive 1 1 1 , 170- 1 71 , 251
article 3 1
aspect 37, 1 33- 1 34
Auslaut 31 , 32, 175
autograph 30, 234
benefactive 32, 1 1 1 , 200, 230, 251
brick 41 , 91
brick-stamp 52-53
building inscription 41 , 62, 77, 78, 91 , 230
case system 32, 53, 54
circumstantial clause 1 98, 199
co hortative 175, 252
comitative case 1 35, 171 - 172
compound logogram 74, 77
compound nouns 1 01 , 1 05
compound verbs 67, 67-68, 70, 95, 1 29- 1 30, 1 36, 138, 176- 177
concrete nouns 69, 1 05, 251
cone 77
conjugation 40, 1 39, 202
conjugation-prefx 34, 39-40, 88-89, 90, 1 36, 1 37- 1 38, 1 38, 175, 200, 202-203, 252
ba 1 36, 1 37, 219, 220
bi 133, 1 37, 138, 21 9
1 34, 88, 90, 175, 200
im 200, 202
mu 34, 35, 88
coordination of nouns 51 , 1 85, 1 87- 1 88
309
310
coordination of verbs 61 , 62, 1 88
copula 1 22, 170, 174
Appendix 6
cuneiform sign formation 26-27, 61 -62, 73-74, 89, 1 04, 1 1 8, 1 25, 1 53, 1 74- 175, 1 82,
193, 1 95, 21 2-21 3, 21 3, 223, 224, 247-248, 252
curse-fonula 1 33, 1 42- 1 43, 173, 1 90
dative case 32, 1 60- 1 61 , 25 1
demonstative 1 35
desiderative 1 36, 173, 252
dimensional-prefx 35, 68, 1 76, 200
ablative 177, 1 97
comitative 1 36, 1 38, 170- 1 7 1 , 200
dative 35, 200
locative 82-83, 84, 1 99, 200
locative-teninative 1 04, 21 8
terminative 68, 69, 70
divinity of Ur III kings 1 1 1 , 1 1 3
door socket 95, 97-98, 1 45, 1 88
double genitive 1 52, 21 6
Ebla 77, 1 1 1 , 1 27, 1 88
.ed 1 85- 1 86, 1 87
enclitic copula 1 22, 1 70, 1 77, 198
ergative case 33, 1 04, 1 62, 1 99, 203, 227
ergativity 33, 1 40- 1 41 , 202f
foundation deposit 9 1
function of texts 41 , 91 , 97-98, 1 88, 254
gender 27, 3 1
genitive 3 1 , 33, 39, 5 1 -52, 54, 54-55, 76, 1 1 1 , 1 21 , 123, 1 52, 229
gunu 1 04, 193, 252
bamt! 37, 40, 1 33
lbbi-Sin 23 1 , 242
indicative 34
intransitive 197, 1 98, 202
loan words 74, 77, 80, 82, 83, 1 03, 1 06, 1 24, 1 27, 1 58, 175, 21 2, 21 3, 21 9, 228, 247
locative case 82, 84, 1 35
locative-terminative case 1 03- 1 04, 21 8, 21 9, 235, 25 1 , 25 1 -252
maru 37, 1 33- 1 34, 1 39f, 1 7 1 , 172
maru-suffix 1 34, 1 39f
Mesannepadda construction 1 54
modal-prefx 34, 1 36, 1 37, 1 40, 175, 1 85, 252
negation 1 85, 198
nominal phrase 32, 35, 1 60
nominalizer 50, 82, 88, 97, 208
numbers 1 09, 23 1 , 232, 233
Index
optional pronominal-prefx 1 71 , 1 72, 176
overhanging vowel 95, 1 23- 1 24, 1 26, 1 52, 173
palaeography 49, 52, 60, 89, 1 1 6, 1 49, 227, 234, 256
participle 50, 82, 1 22, 1 54, 1 87, 208
passive 1 05, 1 97- 1 98, 202
personal-affx, pre-verbal root 35, 89, 1 34, 1 41 , 1 72, 173, 1 86, 1 87, 21 8, 252
personal-affx, post-verbal root 36, 1 34, 1 36, 1 41 , 252
plural - nouns 76, 82, 21 7
plural - verbs 1 72
possessive-sufx 32, 97, 103, 235-236
proper name fonation 27, 33, 1 22, 1 22- 1 23, 1 70, 1 77, 1 86, 21 7, 236, 246
reduced relative clause 1 5 1 - 1 52, 1 54
reduplication 1 39, 1 41
relative clause 88, 1 32- 1 33, 1 5 1 , 1 85, 1 86, 208, 209
restoration of text 207, 234, 246, 250
root 36, 1 36, 198, 200, 252
seal inscriptions 235, 236f, 240
SeSSig 252
Shulgi 1 1 4- 1 1 5, 1 54, 232
Shu-Sin 209, 220, 230, 242
standard inscription 91 , 1 1 2, 1 52
subordinate clause 1 32
substrate 1 06, 1 53, 1 58
syllabic structure 33, 1 21 -2
syllabic writings 252-253
temporal clause 1 32- 1 33, 21 7
tense 37, 1 33- 1 34, 1 85
terminative case 67, 69, 1 03, 21 7
titulary 44-45, 56, 1 1 3- 1 1 4, 1 61 , 1 63, 1 97, 208
! 1 85
Ur-Nammu 44, 56, 1 1 3, 1 1 4
verbal phrase 34, 37, 38
vocalic assimilation 76, 1 24- 1 25, 223, 228
vocative 235, 25 1
votive inscription 70-7 1 , 1 42, 254
votive objects 70, 1 1 7, 1 42- 1 43, 1 77, 21 0, 221 , 240, 254
weights 1 1 2, 23 1 , 232, 232-233, 243
word order 38, 52, 1 61 , 21 7, 21 8, 230, 241
word order typology 40-41 , 55, 1 1 1 - 1 1 2
year-dates 21.9-220, 221
3 1 1
Plhlirafi(ns oflI.1fAS
nu' Infeflltionol lmtituf( for .\fl'soporamil1 Ira Sflldics
TPR I
TPR 1
TPR J
TPR .
TlR s
TPR 6
TPR 7
TlR 8
TPR 9
PUBLICATIONS OF THE TERQA ARCHAEOLOGICAL PROJECT
Modulr Preli miWar)' Reprts
G. BUCCELLATI and M. KELLYBUCCELLATI
General Introduction and the Strarigraphic Record of the Firsl Two Se,ISOIlS
Syru-lfelopowmi ll Sfudies 1/3 (1977).
G. BUCCELLA TI
A Cuneiform Tablet from !he Second Season. Syro.lf.wpowmwn Swdies 1/4 (1977).
M. KELLY-BUCCELLATI and L. MOUNTWILUAMS
Object Typology of Ihe Second Seson (Excluding Vesls ,lnd Lithics)
SyrrMewpatomill Sflldif'. 1/5 (1917).
M. KELLY-BUCCELLATI and W. R. SHELBY
OlfOmic Ves>el TYpolog of the First and Second Seosolls, S)'ro-Meoporomf al/ SllI/ies 1/6 (1 977).
A. MAHMOUD
Die Induslrie der islamischen KeraOli k ous der lweilen Season
Syrf-MeWl/lofOmill Sfudi,, 2/5 (1 978).
G. BUCCELLATI and M. KELLY-BVCCELLATI
Chronicle and Srrlilphk Record of Ihe Third Season. Syro.llesopotollkll Swdill 2/6 tl 978).
O. ROUAULT
Les Documents epigraphiques de la Iroisieme SIison, Syro-Mf''opotollill SmdieI 2/7 ( 1979),
L. MOUNTWILLiAMS
Objecl Typology of the Third Season: The Third and Second Millenni a
S.l'rl.M"SIfJflomil SfJes 3/2 (1 980).
E, GRIFFIN and W. R. SIiELBY
Cramic Vesoel Typology of the Third Season, s.1ru-JlfsIJpmomilJ Swdi<s (forthcoming).
IPR 10 G. BUCCELLATI I,ilh Ihe collnborlion of M. KELLY-BUCCELLATI and J, KNUDSTAD
The Fourth Sea_oil: Introduction and the Slrlillphic Record
Bibli/lhera Mesol}tomira Vot. 10 tl979).
TPR 1 1 N. M. MAGALOUSIS, N al.
Soureing Terhllique for Cermics and Soils. SITO-.lleso{lotomilJ Srudie. 3/' (1 910),
IPR 1 2 G. BUCCElLATI Hlld 0, ROUAULT
Diital PJolling of Archlltological Floor Plln
CAR.WS (CulII/lllrr .it k( ReSfrcl in Nl'Or Eu.'ft'fII Stllllts) 1/1 t l 983).
FinJ Reports
TFK I o ROlIAl!IT, I 'Arhi>e dt PUurum. Bib/it/fflua Mewp/lamio Vol. 16 ( 1 9"4).
R I \'II"S \IJ)S ,\NB RISb\RCH TOOLS IN ANClfNT NEAR I \SUR; STlIOlS
U SIUNNUM. 1 llati/"lpti.n Inde\ of "'ouu"rr ('<lICU Diclimrry of "'-ddf" f/I"plin (I 977)
J 11 SNf-ll A WorLbuul uf (uneifonn Sian ( 1 919)
.
J M SASSON. 'lIled Tnh rrom MITi: A TabuJalion ( (9NO)
IWU l1BL!f4!HI^'
F O. oA 97, Malibu, lalfurna. U.SX 9J5

Potrebbero piacerti anche